If psychology wants the status and rewards of being considered a legitimate science, it needs to make dramatic changes. In the meantime any initial result psychological research produces must be considered not just preliminary, but suspect.
[1] https://en.m.wikipedia.org/wiki/Power_posing [2] https://science.sciencemag.org/content/348/6239/1100.2
So what social-psychological experiments are bogus?
The most famous problem experiment I can think of is Zimbardo's fake prison, and IIRC the primary objection was that the effect was too strong. He let it go on too long and everyone was disgusted.
Rosenhan's experiment is bogus, yes, but I have to wonder... psychiatrists were condoning lobotomies just 15 years earlier. Is it likely that an entire discipline could turn itself around in that time? (Which doesn't excuse Rosenhan, of course)
I am 100% open to your thesis but would like more data.