You've inverted the burden of proof. If someone dreams up a new kind of therapy, it's their job to prove that it actually works; they can't just assert that it

You've inverted the burden of proof. If someone dreams up a new kind of therapy, it's their job to prove that it actually works; they can't just assert that it works based on anecdote. I'm flattering those relatively-untested therapies by assuming that they're all equally effective. In any case, the effect size over pill placebo is extremely small (again, with the honourable exception of CBT for anxiety disorders).