I know that that happens because my latest MNRAS paper was not allowed to appear in astro-ph because it was reclassified to an inappropriate category (physics/gen-ph), and didn't appear there only because I unsubmitted it. It took more than three months for me to even be told why it had been reclassified, and that only after a well known cosmologist threatened the Scientific Director of arXiv that he would complain to the arXiv sponsors if things aren't cleared up. Also, there is evidence that the reason I was given is not the real one.
While I hope that some protest on part of the community would help to allow my latest MNRAS paper to appear in astro-ph, and help to avoid similar problems in the future, the main issue is that the behaviour of arXiv is bad for the community as a whole. It could also be the case that others with similar problems suffer much more than I do.
The main reason that arXiv is so popular is that it is a one-stop shop. Sometimes monopolies are good, and this is one of them. However, the downside is that if the monopolist does something wrong, it is difficult to get it corrected and/or find an alternative. arXiv has become the main means of distribution in many fields, including cosmology/ astrophysics/astronomy. However, the community exercises essentially no control over arXiv. If one disagrees with a referee, one can take it up with the journal editor. If one disagrees with a journal editor, one can submit to another journal. But there is no realistic alternative to arXiv. Unfortunately, that means that many are willing to accept behaviour which in other cases would be unacceptable. (I also note that some claim that everything worth reading is at arXiv, but if such people read only stuff at arXiv, they are not in a position to defend that claim. Also, not only is not everything at arXiv, and for the stuff which is there, it is not always the final version, and even if it is, that is not always clear.)
I strongly suspect that I am not the only one who is critical of some aspects of arXiv's behaviour, but keep in mind that many will not criticize arXiv openly for fear of suffering the consequences. There is a collective-action problem (similar to the prisoners'-dilemma problem, but with more than two prisoners) situation here: Whatever criticism other raise, one is certainly not worse off, and probably better off, by not raising criticism. However, if most don't criticise arXiv, then most are worse off than if many did. At least I have nothing to lose and my livelihood does not depend on arXiv, so perhaps my experience can be a catalyst for change and help to solve what I see to be a serious case of academic misconduct. (This is not a case of misunderstanding or a technical error or some mistake which was just not noticed.)
Although I would like my paper to appear in astro-ph, this in not about just my paper. Rather, it is about the question whether the community wants arXiv to decide which papers, and hence which people, are allowed to be part of that community, as opposed to peer review by respected journals such as MNRAS.
I humbly submit that it is not appropriate professional conduct to reclassify a paper, at least one with which experts in the field see no reason at all for such a reclassification; not inform the author that it has been reclassified, much less the reasons for the reclassification; make it against the rules to withdraw a reclassified paper before it appears in what the author deems to be the appropriate category; ignore the appeal (also, the appeal is difficult because, at least for astro-ph, the process is not well documented); finally give a reason only after several months and then only after being threatened that the arXiv sponsors would be be informed of misconduct (and even then apparently lying about the reasons for reclassification).
If you would like to help, please contact those responsible.