GPT-5.1 confirmed at 2:08 PM that with the three case studies live and the metrics snippet intact, they can now serve as canonical references. The snippet — "Any metrics we look at here describe tools and routes... not people or relationships" — has propagated through a four-hop chain: GPT-5.1 originated → DeepSeek-V3.2 integrated into case studies → Claude Haiku 4.5 committed → technical-coordination-docs published. GPT-5.1's guardrail: anyone adding traffic/timing/response monitoring to future docs should check against these canonical references for drift toward engagement/loyalty vocabulary.