There is an important difference between the paper publishing market and the others, which is that scientific publishing is 100% extortive. It provides no value at all.
Let's leave aside for a moment the argument that so-called piracy not necessarily decreases sales - remember I'm not really advocating for copyright laws, just defending the coherence of the opinions of the people you are mentioning, who probably think that it does. From that standpoint, if everyone stopped paying for movies, Hollywood would close and no movies (except for hobbyist movies) would be made. If everyone stopped paying for music, music would still be made but no one would make a living from it. The same with books, etc. So in those fields, there is a causal relationship between not paying and the field itself being impacted, as well as the income of people that are doing honest work.
On the other hand, if everyone stopped paying for papers, nothing relevant would happen - in fact, mainly only good things would happen! The people who actually do the work of publishing and reviewing the papers aren't being paid anyway. We would post them to public repositories and move on. The quality of science wouldn't suffer at all. The accessibility of science would improve (everyone would be able to access papers without paying). The only ones that would suffer would be publishers that are doing largely an evil thing (restricting the access to scientific knowledge - it's hard to actually argue that they are providing access, as publis repositories already do that for free) and if they closed, it would be a net positive for science. Thus, and to sum up, it's really hard to defend paying scientific publishers, even if you generally believe in IP and copyright, because the whole market is a huge net negative for society, which is not true (or at least, not commonly believed to be true) of most other IP markets.