> So take a moment and ask yourself why any of these publishers shouldn't take the same flak?
It's reasonable to focus on one example of a general problem and try to solve the sub-problem first.
> Of course, the customers complain about the price, so the question is, why not just buy something cheaper? Because that would defeat the whole point, which is to buy prestige. You can't just pay a kid in high school to run a web server for PDFs and say you have a prestigious journal. You have to have a brand that carries prestige, like that Louis Vouitton bag.
This is conflating two things. A Louis Vuitton bag is a Veblen good because people pay more for it because paying more for it demonstrates that they have a lot of money. Publishing in an Elsevier journal is not a Veblen good, because the researcher wants the prestige of producing research of sufficient quality to appear in an Elsevier journal. They don't want the prestige of paying a high publication fee.
Anyway, I'm a computer science researcher. Every paper I've ever written is on Arxiv, and most of them also appeared in archival conferences. Paywall journals are an anti-signal for paper quality in my field. I wrote one paper that appeared in a journal, and that publishing/editing process was nothing but a waste of time to get the material into the journal's format. The publisher added nothing. The continued existence of publishers is a historical accident specific to fields that started when they served a purpose (distributing a hard-copy journal) and have lacked respected senior researchers with enough entrepreneurial vim to set up their own open-access alternatives.
The reviewing process and name recognition is the thing keeping those publishers alive. None of that is provided by the publisher, it's provided by the researchers who work inside it. Earlier I wrote "the researcher wants the prestige of producing research of sufficient quality to appear in an Elsevier journal". That's not really true, they want "the prestige of producing research of sufficient quality to [be accepted by the kinds of reviewers who would reviewer for an] Elsevier journal". ML research has sort of solved this problem via conferences, which are imperfect, but seem substantially better than the journal review process I can see my peers in other fields navigating.
Free markets are cool and all, but sometimes they screw up, especially when they interact with governments and institutions. Publishers currently live in this sweet spot, but science really doesn't need them to.