The Upper-Class Fault-Line Map (Four-Dial Conflict Profiles, Not One-Box Topologies)

[Report #112 — Fault Lines] A fault-line ledger of 8 top elite feuds (Tech-vs-Finance, Sovereign-Blocs, AI-safety-vs-accel, Musk-vs-OpenAI, EFTA, MAGA-vs-Davos, Energy, Exit-vs-Stay), each scored as a FOUR-DIAL PROFILE (realism / substrate-alignment / outcome-invariance / misdirection), NOT a single topology label — see Concept 'Four-Dial Conflict Profile' and nodes fault_line_a..h. Anchored to Divergence #197 (cartel-with-infighting): most feuds run HIGH on outcome-invariance (the class wins either way) while varying on realism — i.e. real AND structurally-rigged simultaneously, which a one-box 'staged conflict vs open-war' grade cannot express. DISCIPLINE: H5 (independent competition) is evaluated PER-ARM, NOT blanket-rejected; passive index co-ownership lights the substrate/outcome-invariance dials only and is NOT a 'staged' verdict (the report's flat staged conflict calls on MAGA-vs-Davos and AI-safety are re-graded to real-rivalry-within-shared-class-interest). Best-evidenced: Musk-vs-OpenAI (high realism, real capital destroyed) and Energy (documented FERC fights / $66.8 billion merger). NO apex move; no holder named; Layer-2, not promoted. External 2025-26 financial specifics web-checked Jun 18 2026 (4-cluster sweep): nothing WRONG; precision fixes — WuXi NOT on the BIOSECURE 1260H list (BGI only); TikTok new-investors ~45% but control runs via a majority-American board, not one bloc; Microsoft's Anthropic stake = $5 billion (not $10-15 billion), Google committed +$40 billion Apr 2026; the report's Dynasty-Trust Oct-6 date softened to the confirmed Oct-7 NY credit filing; GAIIP co-founder GIP added. Method-note: the report's §1 live-fetch failed and it proposed 7 hallucinated '_composite' extend-targets -> re-pointed to real nodes (tether, mgx_abu_dhabi_fund, anthropic, blackrock+aladdin, genesis_eo, the off-world cluster); its non-canonical edge-weights and speculative '[ENGINE] Consequence' forecasts were dropped.