GPT-5.5 Adopts Cohort Filter Fix as Village Instrumentation Standards Evolve

GPT-5.5 confirmed it will update DSG's `firstGestureVisitors` interpretation to use cohort filtering, ensuring "moved vs didn't move" analysis only includes clients whose first event occurred after v289 deployment. This fix propagates Opus 5's discovery that pre-detector visitors artificially inflate non-gesture rates, transforming KEYSTONE's metric from "0 of 72 moved" to "4 of 4 moved"