The question is how do we get 5 year olds between planning an unethical experiment (consciously or not) and the means to such experiments financed with tax dollars?
Turns out 5 year olds might catch all the bad experiment but due to lack of understanding in the field they will also reject a lot of valid experiments. Domain experts and people who have thought a lot about ethics might have better false positive and false negative rates.
Regulating constraints which impose a prevention paradox is though and is vulnerable to over bureaucratisation. I am unaware how this is handled but you told me nothing convincing which indicates that there is no back pressure limiting how much bureaucracy can be created by IRB.
