Comment by veddox

  > From many of these examples, a common thread is the following or a combination thereof:

    - Someone is seen as an outsider within a field

    - The data that supports or undermines claims is noisy. This I think is the core underlying reason, the root of most evil. In turn, the reason for this is lack of tooling to produce precise observations,  or funding to increase sample sizes.

    - Being right, for the wrong reasons. The reasons may be shot down. Rejecting "A->B" does not mean "B" is false, it just means "A"was not a good reason for B.

I think it's important to note that these are actually pretty valid reasons to be sceptical of a claim. (Well, at least the second and third are, but even the first is generally a useful heuristic.)

So I think the issue is not so much that peer-review is broken and needs to be fixed, as that it is by nature conservative and could do with a supplementary process to reduce the false-negatives.