It's entirely self-consistent to believe that the FDA shouldn't be a gatekeeper over what people can try, and also belive that as long as they maintain that gatekeeper role and don't just approve everything non-toxic then it's harmful for them to approve one specific drug that has no effect. If they purport to be a trusted signal of what works and what doesn't, then their approval is an endorsement, and in this case that endorsement is passing off noise as signal.