Cold fusion is quite a good modern example [1]. In that instance peer review was abandoned in an attempt to provide proof by media. The obvious flaws in the science were not made apparent in the nice quiet anonymous fashion that occurs with peer review. Publication did eventually occur as a preliminary note in the Journal of Electroanalytical Chemistry, This was followed up with many critical papers, including one in Nature.
Science is quite resilient. It actually expects theories to be disproved and overtaken by others over a period of time. There is some resistance to change, but when a theory falls it does so quickly. Quite often the search for flaws within a theory yield the most promising insight [2]. However, I suspect theories you are thinking of have a more robust basis than you expect. Some of my favourite theories that have been overturned are as follows:
- phlogiston (an element that supports fire) [3]
- caloric theory of heat (heat is transferred by a specific type of particle) [4]
- plum pudding model of atoms (pretty much as it is stated, see J. J. Thomson) [5]
Science survives all such revolutions where old theories are overturned and new are adopted, largely by rewriting the old theories from the perspective of the new. It gives the impression of continuous progress when the reality is more a progression of leaps and starts.