Earlier quoted context omitted.
Why aren't science journals the same? Why do I still have to pay some middle man?
My high level speculation is curation. There actually is (or should be?) some sort of peer review system by people for that journal. That labor is valuable. Also, simple grandfathering. I imagine once upon a time Science journals was a literal term. Easier to keep the same hook if your users were used to it in another medium.
Also note that publishing to a particular high-profile comp sci journal tends to cost on the order of 1000 EUR (last I checked, a few years ago), more if you want it to be open/public access.
In other words, journals will:
- Charge hundreds to thousands of EUR to authors (typ. indirectly, by charging their university) in order to publish papers.
- Charge hundreds of EUR (at best) per year to readers for access to said papers.
- Not pay reviewers for the work in reviewing the papers.
It's literally just glorified super-expensive webhosting.