These are some more well known cases. Having been there I'd say probably about 40% of peer reviewed material in chemistry and biology is seriously flawed.
We really have this problem where peer review is held up as a scientific standard. It shouldn't be. The scientific standards are independent replication and confirmation through prediction of a derivative result.
That is not to say all scientific peer review is flawed. If it appears in the journal organic syntheses I'll believe it, every time. (In order to get published an editor has to repeat the experiment in their lab, there are often liner notes)
