Wrong as a general rule, although there are some specific situations that mandate RCTs.
Not necessarily.
Firstly, fascinatingly enough,
there is a Cochrane review analysing the difference between effects measured by observational studies and RCTs in and finds they are... very small. Observational studies can be powerful and accurate, and should not be casually neglected.
Broadly, you would expect that a RCT should be better than an observational study
on average (even if the above meta-analysis did not find much difference between the two). But that's not always the case on an individual level, because there's a whole welter of factors that go into the quality of a study. Or simply, if you like, a good observational study is better than a bad RCT. That's one reason why you can't always get too hung up on RCTs for everything, and certainly cannot just assume a RCT is superior.
A delve into the literature would find many, many experts noting that it is exceptionally hard to do a RCT on masking to reduce respiratory infection, both for practical and ethical reasons. What that also means is that RCTs done for masking are often likely to have substantial flaws and limitations. Therefore a meta-analysis on a bunch of relatively low-middling quality RCTs is inherently limited in its conclusions, because you can't turn lead into gold. And that Cochrane study - because Cochrane is actually good at this - essentially says little confidence can be held in its own result. So, in fact, we might actually get
better information from observational studies than RCTs in such a situation: RCTs are just the wrong tool for the job.
And that's why I have a problem with certain over-zealous adherents of RCTs. They've taken a general principle and turned it into an absolute rule. That then causes them to undervalue/ignore potentially useful data... and undervaluing/ignoring good data is bad science.