Peer review isn't "broken", it's just working badly because no one has time to do it properly. Yes, publishing pressure is most of it. That combined with the difficulty of changing subject and a strong survivor bias.

By "survivor bias" I mean that the people who hold positions in academia are the ones who are comfortable with how the system works. They see nothing wrong with publishing on topics they don't believe will contribute to progress anyway. It's just a job.

They will often rationalize this by saying that they might find something interesting while working on this other thing that they get paid for. This is a very wide-spread tale in the foundations of physics. Never seems to happen though, funny, no?

And once you've spent several years on a topic it's difficult to start something new. Even if you're tenured, these days your funding for other people might depend on you continuing the same thing you've been doing already. The path of least resistance is to continue.

This is why academia has an incredible inertia, it takes decades to correct mistakes, and there's a risk that corrections don't happen at all. Most scientists trust blindly in "self-correction" but don't seem to realize that it's them who have to do that correction.