> Is that enough to justify the conclusion that “mechanics are bad”?

Maybe not, but it's certainly enough to justify the conclusion that "mechanics should not be blindly trusted". Same with scientists. Doing good science is hard, and it wouldn't surprise me at all if most studies are flawed.

Also, there's a reason the concept of independent verification is so important in science. Even if the scientists doing a study are trying their best to do good science, it's incredibly easy to make a mistake. A healthy dose of skepticism is very much warranted.

You shouldn’t blindly believe anything, but science, faulty as it is, does entail a higher degree of credibility than, say, anecdotal evidence.

Perhaps, but that's not a very high bar. If you personally experience something that disagrees with what a research paper says, I don't think it's reasonable to just assume you must have been mistaken.

That doesn't mean I think you should believe your senses either. It means it's worth looking into it further (if you care, at least).