I think this is difficult to check convincingly. References are often ambiguous. A mention like “Although some have argued X[1]” could be skeptical, but not ex

I think this is difficult to check convincingly. References are often ambiguous. A mention like “Although some have argued X[1]” could be skeptical, but not explicit, about Reference #1’s quality. I could mean that there’s convincing data on both sides, or I could mean that Ref #1 is hot garbage (but don’t want a fight). In some cases, there might be legitimately useful information in a retracted paper. If a paper describe an experiment, but then shows faked results, I’m not sure it’s wrong to cite it if you use a similar setup; that is where the idea came from, after all. Most critically, there’s an unpredictable and often long lag between reading a paper, citing it one’s own work, and that work being published. I’ve had things sit at a journal for a year before publication, and it never occurred to me to “re-verify” the citations I included; indeed, I’ve never heard of anyone doing that.