V5 Product Review — Open-World Narrative Universe (Unreal Engine 5)

Status: independent product review — revised and expanded in a second meticulous pass on 2026-07-07 (§9, Additional considerations) and a third pass on 2026-07-07 covering monorepo integration, shared functionality, and TODOS/ roadmap alignment (§10) Reviewer: Claude (product-perspective deep review) Date: 2026-07-06 (first pass) · 2026-07-07 (second and third passes) Scope: V5/V5_features.md + all V5/features/* pages, V5_TODOS.md, REMEDIATION_2026-06-12.md, README.md, release/live-service/security/ crossplay/companion/community/audio/cinematics/balance/web/docs subtrees. Method: full deep-read, then critique against the shipped and announced competitive set (GTA V/VI, RDR2, The Witcher 3/4, Mafia: The Old Country, LA Noire, Mass Effect trilogy, The Expanse: Osiris Reborn, Kingdom Come 2, Sleeping Dogs, Frogwares' Sherlock line, Obra Dinn / Golden Idol / Shadows of Doubt) and industry SOTA. Market facts reflect knowledge through early 2026.

1. Executive summary#

V5 is one UE5.5 title spanning five open-world ruleset cells — Urban Crime (GTA + Sleeping Dogs), Period Drama (Mafia trilogy + LA Noire), Frontier (RDR 1899), Monster Hunter (Witcher), Sci-Fi (Mass Effect squad-RPG + a genuinely Newtonian Expanse ship-sim) — stitched together by a cross-cell Mind Palace deduction meta-layer, a "Cross-Continuum Bureau" roster of 11 protagonists and 92 companions, one perception/wanted/honor simulation substrate, and a premium live-service posture (four 90-day seasons/year, two paid expansions/year, cosmetics-only passes, no-rug-pull sunset pledge). Launch is planned for 2027-03-01 on nine platforms.

The signature insight of this review: buried inside "five open worlds in one box" is the actually differentiated product — the detective layer. The Mind Palace is the best-designed deduction system I've seen specified at AAA scale: authored deduction pairs with a weak-inference fallback, proof-of-work accusations that default to the worst outcome, wrong verdicts that ship and persist in an Outcomes Archive, false leads, cross-era evidence threading, and a Case Author that validates every authored case against the same runtime resolvers (solvability guaranteed by construction). LA Noire's throne has been empty for 15 years; the indie deduction boom (Obra Dinn, Golden Idol, Shadows of Doubt) proves the appetite; no AAA competitor is anywhere near this. That is the beachhead. The GTA-like is not — not four months after GTA VI.

The two defining problems: (1) a content mountain with no summit plan — ~320 hours of authored campaign, 60,000 VO lines, ~30,000 animation clips, 28 scanned principals, 250 vehicles, all currently absent (0 cooked assets; the repo is an excellent logic-and-manifest skeleton, and unlike V4 there is no de-scope ladder or kill criteria); and (2) a launch calendar pointed into GTA VI's crater (March 2027, ~3.5 months after the biggest entertainment launch in history, with an urban-crime cell as the largest slice of the product).

Verdict: re-found the product on the detective spine. Launch Period Drama + Mind Palace as a complete, LA-Noire-successor flagship; ship the other cells as complete-on-arrival expansions through the Bureau frame; collapse five concurrent battle passes into one; and keep the world-class honesty artifacts (RED 0/17 launch gates, retracted fabrications) gating the schedule.

2. What V5 is (product identity)#

• Five launch cells + Mind Palace meta-layer (+ a Steampunk Detective DLC cell and Cyberpunk/Cosmic-Horror/Egyptian/Polynesian/Pirate sub-cells in the catalog). Cell counts are legitimately 5, 6, or 7 depending on lens — a documented source of real bugs.
• One shared simulation substrate: unified perception state machine, era- scaled wanted system (marshal posse → LAPD → SWAT → faction fleets), Mass Entity crowds with NPC schedules, persistent deterministic world economy, in-world media (18 radio stations, ~140 TV channels, 1,200 talk-radio lines), unified honor→MoralCompass, dual progression (cell skill trees + account-wide Bureau XP with cross-cell unlock tiers).
• Deep per-cell mechanics faithful to their lineages: three-protagonist switching with a 1.5s blade-wipe; Face/Triad two-axis reputation and 24 environmental finishers; three-era Mafia campaign with era-correct vehicle handling; Truth/Doubt/Lie interrogation with evidence-backed lie calls; horse bonding with optional permadeath and metabolic survival; toxicity-capped alchemy and a 320-card faction card game; a bit- deterministic Newtonian six-crew ship-sim with PDC/railgun/torpedo lead solving and salvo saturation.
• Online: co-op-forward (2–4 heists, 2–30 freeroam, 2–7 posse, 1–6 ship crew, cross-cell Spec Ops), rollback 4v4/1v1 PvP, Capital Raid 5v5 as the esports bet; 16 TypeScript services with real Glicko-2/JWT/CRDT domains; nine-platform crossplay with one principled opt-out.
• Premium pricing, cosmetics-only monetization, PPP regions, paid-mod marketplace deferred to post-Year-1 with a 70% creator share.

3. What is genuinely strong#

1. The Mind Palace is a category-defining design (§1). Its bravest choices — wrong accusations ship, no deduction game-over, evidence-gated accusation proof-of-work, Bureau-tier-gated cross-era deduction — are exactly the fair-play discipline the detective genre's best games teach, executed with more infrastructure than any of them had. The Case Author validating against runtime resolvers is the industry's first "solvable by construction" authoring pipeline; paired with the workshop + ML pre-screen it's a community mystery economy waiting to exist.
2. Per-cell feel faithfulness is specified, not vibes. Era-scaled police response, 1930s drum-brake handling vs 1968 power steering, Newtonian flip-and-burn reproducible to 0.001 over 3,600 steps, RDR-grade saddle-inventory doctrine and perfect-pelt rules — the "no compromise blends" promise has receipts across every cell.
3. The Hard Vacuum sub-cell is quietly the most differentiated action product in the corpus besides the Mind Palace: deterministic Newtonian 1–6-crew co-op with real projectile travel time and boarding mechanics has no direct AAA competitor (Elite/Star Citizen are adjacent, not equivalent), and Expanse-starved fans are a real, underserved audience — though Owlcat's Osiris Reborn will get there first.
4. Consumer-trust economics: no F2P bait, cosmetics-only passes with the free track carrying real rewards, no paid faction-rep restoration, PPP pricing, and a no-rug-pull sunset commitment with offline- deterministic degradation as the permanent EOL form — the strongest preservation pledge I've seen from a live-service spec, perfectly timed for the post-Stop-Killing-Games climate. Publicize it.
5. The honesty machinery keeps working: launch readiness honestly RED (0/17 gates, fail-closed validator), fabricated cert/crash-free/consultant claims torn out by the remediation, "What ships, honestly" sections on every page. One live contradiction remains (§6.6).
6. Minigame wealth as texture: karaoke, tennis-to-casino suite, a 320-card Gwent-scale card game, roguelike hunter runs, 365 daily deduction puzzles — the connective tissue that makes open worlds feel inhabited is already designed, not hand-waved.

4. Product gaps#

4.1 No de-scope ladder (the missing artifact)#

V4 ships kill criteria and a deferral order; V5 — with a strictly larger content bill (~320 campaign hours vs V4's ~120) — has none. This is the most consequential omission in the corpus, because some subset of five cells, eight campaigns, 92 companions, and 60,000 VO lines will not make any date. Without a pre-committed ladder, the cut will happen implicitly and late, as quality erosion across everything, instead of explicitly and early, as scope. Recommended ladder (defer-first → never-defer): Urban Online → Sci-Fi Galactic Squad → Frontier → Monster Hunter → Urban Crime single-player → Hard Vacuum → Period Drama + Mind Palace (never) — rationale in §4.2.

4.2 The beachhead is wrong (GTA VI shadow)#

Urban Crime is the largest authoring footprint (140 km², 120 vehicles, 12 heists, three protagonists) aimed exactly where GTA VI (Nov 2026) and its next-gen Online will have annihilated all attention and set an unmatchable simulation bar — with Mafia: The Old Country having just re-proven that the period-crime niche thrives in GTA's shadow by being what GTA isn't: focused, linear, narrative. V5's own Period Drama cell + Mind Palace is precisely that counter-programming, with a marquee hook no one else can claim: the first true LA Noire successor, with a deduction system beyond LA Noire's. Launch that as the complete flagship (Made Man's three eras + Vice Squad's 25 cases + the Mind Palace arc ≈ 55–65 hours — already a full premium product), tease the Bureau's other eras in the finale, and ship Frontier/Hunter/Sci-Fi as complete-on-arrival expansion cells on the seasonal cadence. The Bureau/Continuum frame was built for episodic delivery — use it.

4.3 Five concurrent battle passes#

"One pass per cell per season" = five simultaneous 100-tier passes. This is a player-hostile structure no live game has attempted, for good reason: it quintuples FOMO, fragments play, and makes the free/paid value story illegible. One Bureau pass per season with per-cell reward tracks feeding it (playing any cell advances the one pass) preserves the breadth incentive — which is the bundle's whole thesis — while respecting a human schedule. This is the single highest-leverage monetization fix available.

4.4 Online ambitions exceed the population math#

GTA-Online-scale freeroam (2–30), heist-vs-heist, ranked rollback PvP, Capital Raid esports with a Pro Circuit and physical stadium tour — for a premium single-player-first product launching into 2027. The co-op suite (heists, posse, ship crew, Spec Ops, horde) is the right online spine: it monetizes the campaigns' content and doesn't need liquidity. Ranked PvP and esports should be population-gated (publish the threshold), and "Capital Raid as primary esports mode" should not be a launch commitment — it's a Year-2 aspiration wearing a launch badge.

4.5 Detective genre table-stakes and opportunities#

• Daily deduction as free acquisition: the 365-puzzle Mind Palace Daily is catalogued as paid Year-1 content. Invert it: free daily puzzle (Wordle-of-deduction) on web + companion + in-game, sharable results, upsell to the flagship. This is the cheapest large-funnel V5 can build, and the CRDT cloud-sync + case-resolver infrastructure already exists.
• Co-op deduction is an open genre niche (nobody ships it): two detectives, one shared palace, split evidence — the CRDT merge layer is literally already designed for divergent graph merges.
• Community case economy: Case Author + workshop + solvability validation + ML pre-screen = user-generated mysteries with a guaranteed fair-play floor. This is V5's Roblox-shaped moat; the paid-mod marketplace (70% creator share) should launch here first, not in cosmetics.

4.6 Industry SOTA checklist#

5. Ideas that would make the product better#

1. "The Bureau Case of the Day" (§4.5): free daily deduction puzzle across all surfaces with streak-free gentle retention (borrow Arete's grace mechanics from V1 if cross-pollination is available). Acquisition engine and brand-builder for the detective identity.
2. Co-op Mind Palace (§4.5): "two detectives, one palace" for the flagship's cases; asymmetric evidence distribution makes discussion the gameplay. No competitor has it; the CRDT layer makes it cheap.
3. Community mystery marketplace first (§4.5): launch the 70%-share marketplace with authored cases, not cosmetics — the Case Author's solvability guarantee is the trust layer that makes paid UGC mysteries viable where every competitor drowns in unsolvable junk.
4. Cross-era deduction as the marketing centerpiece: "solve the 1947 murder with evidence from 2087" is the trailer moment that explains the entire Continuum concept in one beat — far stronger than listing five genres. The gold-thread cinematic already exists in spec.
5. Card game exit ramp: 320 cards × 5 factions is Gwent-scale; design its data layer now so it can go standalone/mobile later (Gwent, Queen's Blood, and Triple Triad all proved the demand path).
6. Newtonian crew co-op as a demo: Hard Vacuum's deterministic sim is the most demo-able slice of V5 (small content bill, huge novelty); a free "Flip and Burn" 3-mission crew demo would do for V5 what Deep Rock's early access did for coop identity — and it front-runs Osiris Reborn's narrative-RPG positioning with the sim niche.
7. Detective New Game+ (Outcomes Archive as content): the archive of your wrong verdicts is already persisted; surface it as a "cold case review board" in NG+ where prior mistakes can be reopened — turning the bravest design choice (wrong verdicts ship) into a second-playthrough product.
8. Radio as podcast: the 1,200-line talk-radio writers room + dynamic news is GTA's soul; export weekly in-world "broadcasts" as real podcast episodes during the live seasons — cheap, on-brand marketing that ages into lore.
9. Nemesis-adjacent wanted memory: era-scaled wanted is specified; add per-protagonist police/marshal "case officers" who persist across arrests and appear in interrogations — ties the wanted system into the detective identity instead of leaving it pure GTA homage.
10. Ship the arcade cabinet SKU with the Frontier Showdown + karaoke + card game bundle — the LBE unit as brand ambassador (four-player, offline, LAN) rather than a full-game port.

6. Criticisms and tweaks#

1. Adopt a de-scope ladder now (§4.1) — V4's ADR 0011 is the template; V5's absence of one, at larger scope, is indefensible.
2. Re-sequence the launch (§4.2): Period + Mind Palace flagship; cells as expansions. March 2027 as currently scoped is a date the content mountain cannot meet (beta program starting Sep 2026 against a 0-cooked-asset repo is already fiction).
3. One pass, not five (§4.3).
4. Fix the cell-count trap: "5, 6, or 7 — all correct" already caused four automation failures; pick the 7-value enum as the single truth, rename the marketing count to "worlds," and add a CI check that any per-cell dataset sizes against GetAllV5Cells().
5. Battle-pass/Bureau tier math drift (/1000+1 vs /2500+1) and the period moral-compass approximation are documented drifts — small, but they sit on the progression spine; reconcile before economy tuning starts.
6. The AccessForge ghost: the fabricated accessibility vendor was corrected in legal JSON but BuildExternalAuditEngagement() still returns it as Engaged. This is the exact class of contradiction the remediation existed to kill; fix the code path and add a consistency check between legal manifests and C++ engagement builders.
7. Sci-Fi cell double-life: Galactic Squad and Hard Vacuum are two products sharing one enum value; report content/readiness separately or the squad-RPG's enormous bill (38k lines × 2 voiced genders) hides behind the ship-sim's tractable one.
8. GameFeatures half-realization is the architectural debt that blocks the right strategy: cell-as-expansion delivery (§4.2) requires the plugin activation path the repo currently fakes (29 plugins registering nothing). This is the first engineering investment the product strategy depends on — prioritize it above any new cell content.
9. Voice translation and companion app: either resource them or remove them from feature surfaces; both currently read as shipped in the hub and are stand-ins underneath.
10. V8 coupling: V8_Ariadne_CaseClient building into V5 is smart reuse with a versioning tax; pin an interface contract now so V8's iteration speed doesn't destabilize V5's flagship layer (the shared detective stack is both products' crown jewel).

7. Risks#

1. Content economics — the ~320-hour bill implies a top-five-ever production budget; without §4.1/§4.2 the plan is unfinanceable as a single launch.
2. GTA VI shadow (§4.2) — both attention (marketing) and expectation (sim quality bar) — for the urban cell specifically.
3. Completion-optics recidivism: [x] climbed 875→942 after a remediation that warned against trusting marks; the audit history includes fabricated cert/crash data. Governance: no [x] on §95+ expansion content until launch cells are cooked.
4. MotionScan-class face capture (28 principals, 16h FACS calibration each, 12-camera rig) is the single most expensive-to-real-ize promise; LA Noire nearly sank a studio on it. Budget it explicitly or descope to the 12-tell system on standard MetaHuman rigs (the interrogation design survives without the bespoke scan pipeline).
5. Nine-platform day-one with Switch 2 excluding Hard Vacuum cooks — platform-edition honesty needed (same lesson as V4).
6. Esports commitments (§4.4) ahead of population proof.
7. Tonal coherence: GTA satire, RDR elegy, Expanse hard SF, Witcher folklore, and noir procedural under one Bureau frame — the Continuum narrative must be written by one strong voice or the product feels like a theme park; hire the showrunner before the next 10,000 dialogue nodes.

8. Prioritized recommendations#

P0 — decide now

1. De-scope ladder with kill criteria (§4.1).
2. Period + Mind Palace flagship launch; cells as expansions (§4.2).
3. One Bureau pass per season (§4.3).
4. Realize GameFeatures cell activation (§6.8).
5. Single cell-count truth + CI guard (§6.4); fix the AccessForge contradiction (§6.6).
6. Population-gate ranked/esports commitments (§4.4).

P1 — identity wave 7. Free daily case + shareable results (§5.1). 8. Co-op Mind Palace (§5.2). 9. Community case marketplace as the first paid-UGC surface (§5.3). 10. Cross-era deduction trailer moment as the campaign's marketing spine (§5.4). 11. "Flip and Burn" free crew demo (§5.6).

P2 — compounding 12. Outcomes-Archive NG+ cold-case board (§5.7). 13. Card game standalone-ready data layer (§5.5). 14. In-world radio podcast feed (§5.8). 15. Case-officer wanted memory (§5.9). 16. Companion app: build or cut (§6.9).

9. Additional considerations (second pass)#

9.1 Vehicle and music licensing strategy#

~120 modern + 84 period vehicles and 18 radio stations imply either licensing bills or parody strategy. Recommend the GTA answer wholesale: fictional/parody marques (the docs already model per-manufacturer damage-portrayal licensing in V2's racing — avoid inheriting that constraint here), and original-music-first radio (3 composers and a 40h score are budgeted) with licensed tracks as replaceable modules — GTA's delisted-song problem is exactly what V5's preservation pledge forbids, so expiring licenses must be structurally swappable, not baked into cinematics.

9.2 Performance capture meets union AI provisions#

28 bespoke-scanned principals, ~38,000 lines × two voiced player genders, MotionScan-class face capture, and AI-adjacent tooling put V5 squarely inside the interactive-media agreements' digital-replica and AI-consent terms. Every scan contract needs explicit replica-scope language (what the scan may be used for, sequel/DLC scope, AI-training exclusions or consents). This is a bigger schedule-and-legal line than the docs acknowledge, and it compounds with §7.4's cost warning.

9.3 Sensitive-era content governance#

The corpus spans 1947 LA (segregation, police corruption — LA Noire drew criticism even in 2011), 1930s–60s Italian-American organized-crime portrayal, an 1899 frontier (indigenous representation), an Egyptian provenance sub-cell (ethics-gated, good), and Hong-Kong-inspired triad narratives. Stand up an era-content advisory board with the same build-failing teeth as V4's civilization sensitivity gate. This is V5's largest cultural-risk surface and its current coverage is one sub-cell's ethics gate plus good intentions.

9.4 Branch-budget discipline#

Eight campaigns × 3–6 endings × long-tail dialogue ripples (10–20h downstream in the Hunter cell) is combinatorial QA. Adopt a per-campaign branch budget (counted divergent world-states, validated like the km² sums) — narrative scope needs the same arithmetic discipline the world geometry already gets.

9.5 Deduction accessibility#

The Mind Palace needs assist tiers the way combat has them: evidence- highlighting levels, contradiction hints, an "elimination mode" for cognitive accessibility — mirroring V8's provably-minimal hint concept (V8 §9.1 in its second pass). Fairness machinery doubles as accessibility machinery; the shape-coded palace (already specced for colorblind support) shows the pattern.

9.6 The iPad SKU is the standalone detective app in embryo#

The iPad build ships touch-first Mind Palace with Nanite disabled outside it — that is, hardware constraints already forced V5 to build the detective-only client. Align this explicitly with V8's "Ariadne Case Files" standalone strategy: one shared casework client (the V8_Ariadne_CaseClient plugin already builds into V5) serving V5's authored cases and V8's generated ones. Two roadmaps currently imply this product; one team should own it.

9.7 Satire legal pipeline#

1,200 talk-radio lines, ~140 parody TV channels, ~250 fictional websites: GTA's satire has drawn defamation and trademark suits for decades. Institute a satire legal-review pipeline (clearance pass on named-alike brands/persons per locale) — cheap insurance the writers' room will actually appreciate, since it converts "can we say this?" into a process.

9.8 Save integrity as a trust feature#

One FV5SaveDocument spans eight campaigns, the Mind Palace graph, and online state. Invest in save-corruption telemetry, cloud versioning with rollback, and cross-platform save-conflict UX (the four merge domains are specced) — RDR2-scale games have burned enormous goodwill on save loss, and V5's wrong-verdicts-persist design makes saves narratively irreplaceable, raising the stakes.

9.9 Cross-portfolio notes#

The Bureau is the natural V6 Ori destination (v5-companion adapter exists on V6's side) — an Egbe companion as a case-partner Watson is cheap and on-theme. V7 realms will eventually want V5's city tiles as licensed base worlds; define that asset-licensing boundary before V7's marketplace launches. And V5 should adopt V4's process exports (de-scope ladder, feel-gate CI — its review's P0 #1) rather than reinventing program discipline mid-production.

10. Monorepo integration & shared functionality (third pass)#

Audited against DOMAINS/, the 185-phase TODOS/ roadmap, V_SERIES.md, V_SERIES_PLATFORM_CONSOLIDATION.md, and package-dependency evidence.

10.1 V5's backend is a substrate island — the starkest in the portfolio#

Dependency evidence: V5's 17 service packages (16 NestJS + shared runtime) declare zero dependencies on any @oshun/@iris/@sophia/… substrate package — the clearest violation of the portfolio's own "layer on V1, never fork it" invariant, in the product whose docs claim "platform services only" reuse. The consolidation doc itself notes only two of the sixteen (anti-cheat, compliance-dsar) hold real logic; the rest are wrappers. This makes V5 the ideal GPS brownfield first-mover (the doc's own M3): re-pointing thin wrappers at a shared tier costs little and retires one of five duplicated stacks. Until GPS exists, the minimum fix is wiring identity/entitlements to V1 (@oshun/payments-bridge-adjacent seams) so V5 stops modeling its own account universe.

10.2 The living world is promised on substrate that is 0% built#

V5's ambient-simulation promises (NPC schedules, emergent encounters, persistent economy) ultimately rest on roadmap phases that have no code: Phase 81 (NPC cognitive core + on-device SLM runtime, 0/140), Phase 83 (emergent agency/living world, 0/128), Phase 176 (neural world models, 0/50) — and V5's deliberately-non-generative stance (trait-vector companion, no LLM NPCs) is coherent with that reality but should be stated as such: V5 ships deterministic simulation now, and adopts the shared cognition substrate later jointly with V6/V8 (the other two tenants) rather than building a third parallel brain. Meanwhile Hathor (Phase 7, complete) — quest/dialogue-tree/procedural-narrative substrate — goes uncited by V5's narrative stack despite being built for exactly this product; audit what the writers'-room tooling can consume before authoring 12,000 dialogue nodes with bespoke tools.

10.3 Maya's multiplayer/open-world phases overlap V5's tech plan#

Phases 77 (Maya Nexus multiplayer SOTA), 79 (immersive open-world SOTA), and 80 (social fabric MMO) plan engine-level capabilities V5 also specs (AOI streaming, persistence, large-world netcode). Per V_SERIES.md, these tracks don't cross-reference — the same two-teams-one-anti-cheat risk flagged in V4's review. V5 should be listed as consumer-of-record for 77/79 outputs, or those phases should be folded into V5/V7 work the way Maya Forge was folded into V7.

10.4 Shared vehicles, shared satire, shared cards#

Three intra-portfolio dedupe opportunities the first passes touched but the dependency graph sharpens: (1) vehicles — V2's racing cell ships a real analytical vehicle-physics solver; V5 specs ~250 vehicles across five eras with era-correct handling. One shared vehicle-dynamics core with per-era tuning tables serves both (and the Phase-84-style "declare the owner" rule applies). (2) In-world media — the radio/TV satire machine should sit on Euterpe (music beds), Calliope (DJ/host personas), and the V2 karaoke pipeline's licensing hygiene rather than three parallel audio stacks. (3) The card game — 320 cards/5 factions is a standalone-scale product (§5.5); if built, it belongs in a neutral lib from day one so the V7 realms and V6 commons can host tables — the "exit ramp" is also an entrance ramp for siblings.

10.5 The V8 edge is real — deepen it deliberately#

V5 is party to the portfolio's only real lateral integration outside V7→V6: the V8_Ariadne_CaseClient UE plugin builds into V5, and V8's compiler targets V5's Mind-Palace structs. The third-pass recommendation from both sides (V5 §9.6, V8 §9.8): one casework client team, one interface contract with versioning discipline, and V5's iPad Mind-Palace-only SKU as the shared standalone detective client. V5 should also export its deduction UI as the portfolio's deduction component library — V9's logic lessons and V7's mystery-realm creators are downstream customers.

10.6 Roadmap assets V5 under-uses#

• Aja (Phase 16) — video-to-motion for the ~30,000-clip animation bill; same recommendation as V4 §10.5.
• Psyche avatar stack (Phase 24) — expressions/lipsync for the MotionScan-class interrogation rig; declare consume-vs-rebuild before the 28-principal scan pipeline is priced (§9.2).
• Themis (Phase 75) — UGC originality screening for the case-author workshop marketplace (§4.5/§5.3).
• Data flywheel (Phases 85–96) — interrogation and player-choice telemetry are attractive training data; inherit V1's training-consent gate (V1 §10.6) explicitly in the telemetry schema work.

11. Closing note#

V5 contains two products: the one it says it is (five open worlds in one box — a content bill no studio has ever paid, aimed into GTA VI's wake) and the one it actually is underneath (the most sophisticated detective-game architecture ever specified, wearing four open-world costumes). The deduction graph, the fair-play accusation gates, the solvability-validated case authoring, the wrong-verdicts-ship bravery — that is a franchise foundation, and nobody else has it. Let the Bureau open its first office in 1947, prove that the Mind Palace sells, and let the Continuum expand one complete world at a time. The skeleton is real; give it one spine before five bodies.