They're both scenarios where we push things to an extreme. As I see it (from a kind of Naturalist+Existentialist-I-Guess perspective based on my reading of Schrödinger's lectures on thermodynamics and more contemporary researchers who have posed existentialist questions) all things have this entropy nature in them, a pull towards disorder (don't read this as chaos), variation and ultimately a sort of dispersion. If we do nothing, everything will crumble, if we do everything we can to stop one thing from crumbling, then everything else will crumble faster: the key is to find the right amount of "push back" needed for the right things, the aim for the optimal instead of the ideal.

So any system we create will need reforms and changes, but there is no formula to "when", "how often" and "how extensively", besides it's something that a different generation needs to deal with anyway and we cannot predict what paradigms and context will be relevant in the future. Providing a framework for the means to make changes and giving a rational permission for occasional radicalism is how systems can overall preserve its function.

The way this relates to the OP is that since we have unprecedented, massive influx of new research publications, there is a lot that just goes kind of unchallenged, this is tied in with entropy, which gets "stronger" with "bigger numbers", so we can work against this by verifying the research in some way, but just how much resources put into this verification is worth when research is just getting more and more expensive? Currently the main form of this is by doing metastudies, which collect a bunch of relevant studies, critically studies their methods, results, impact on other researches, etc. to determine whether this was in some form verified, falsified or is suspiciously lacking. It is essentially doing science on science, and in many fields it seems to work just fine, though typically the criticism is that there isn't enough people doing this with enough resources, which I agree with, but I'd also add, that metastudies are hard to do because of the highly critical focus of it.

Overall, this OP is a bit dramatic, there is no real crisis, in the general scientific community, so no, scientists are not shacks (yet), but in some fields this has grown out of proportion (psychology and medicine namely) where a bit of change is needed or more frequent/sizable metastudies.