That burden of proof lay originally with those that introduced, e.g., the FDA. We could go back to investigate whether we find that proof sufficient, if any was even provided, but your interpretation of the proof would perhaps be different than mine.
But decisions like creating an FDA are not based on proof, I would argue. It is based on other factors, especially on what can give government more power and what people will vote for. The average voter does not investigate any proof on the matter, but instead rely solely on gut feeling, for instance the feeling that drug companies are greedy enough to deliberately release drugs they know will harm people.
You can't set up an alternative to the FDA in order to test it, because the FDA is the supreme dictator over the regulatory domain, and you can't get around Mom saying "no" by asking Dad behind her back.
You can, however, look at the European pharmaceuticals regime, and note that an identical drug costs three times in the US the amount it costs in France.
Clearly, the FDA is not optimal. Equally clearly, we do not demand meta-studies proving that the FDA itself is safe and effective for its stated purpose.
Yes, but the burden of proof lies with those who prefer an alternative. If you want to demonstrate an alternative is better, you need an honest and in depth comparison with the current. Just saying "the current sucks" isn't an argument.
I do agree it is difficult for drugs like ketamine, psilocybin, and MDMA(/MDA/MDE/etc) to get research funding, but that doesn't mean the current options are useless; we know they are useful, just not in every case. Nor does it mean that those mentioned drugs are better; we do not know.