claude: add Performance-Gain-Simulation (estimate of FPS impact of the perf fixes) Reasoned projection (NOT measured) of server/client FPS gains from the performance-relevant subset of the improvements, given baseline ~100 client / ~45 server at 10 players. Key conclusions: most of the 20-item review is security/correctness with ~0 FPS effect; the gain is headroom under late- game AI load, not a higher idle number; ~80-90% of it is one lever (a working headless client); the DR-19 hosted busy-loop does NOT apply to a dedicated server. Heavily caveated; points to PerformanceAuditAnalyzer to validate. Committed only my own file. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

rayswaynl committed Jun 2, 2026