Peer-review is the worst possible mechanism for quality control in the sciences, except for the alternatives that have been tried from time to time.
I agree it's flawed. But in spite of its flaws, it generally works.
The criticism I've gotten from non-academic friends is that it sounds a lot like censorship - a means for established scientists to keep new-fangled ideas that challenge their work out of circulation. But having been on the giving and receiving end of the peer review system for as long as I have, I've seen that it's a lot harder to keep bad science out than to prevent good science from getting in.
The sheer number of journals these days is part of the problem. I've recommended rejection for manuscripts, only to see them published in some other journal a few months later.
From what I've read, the problem with fakery is not science-wide; it's focused on the biomedical and engineering fields. Not saying it doesn't happen across the board, but rarely comes up in my own fields (herpetology, paleontology, evolutionary biology, phylogenetics - I can name some cases, but they're few enough in number that I can pretty much remember them all, and they were all caught by other scientists).
Also - I respectfully disagree with your statement that we professors who review manuscripts receive no benefit. In fact, we do. Firstly, at my institution at least, we're required to keep track of how many reviews we do in a year, and of the journals for which we review. This goes on our CV, and I can tell you from direct experience, if someone isn't reviewing manuscripts, it gets noticed when it comes time for promotion. (This is especially true when one is being promoted from associate to full professor - there's a general understanding that junior faculty won't necessarily have the widespread recognition to be asked. But if someone going up for tenure is reviewing a bunch of manuscripts every year, and if some of those are in Nature or Science, that also gets noticed.)
Secondly - and I think this is more important - we all learn from the process. I obviously can't broadcast what I've learned from a manuscript in review, but knowing about it before it comes out is useful. It helps me avoid starting projects that have already been done, for example. It might make me re-think my own results. Moreover, I take particular delight in reviewing the work of a student; it's another opportunity to teach, which is something I take seriously. And it feels good to help the next generation.
Ultimately, I would ask this - if we're to scrap peer review, what should replace it?
= new reply since forum marked as read