In many cases, papers from 20 or 50 years ago, which were once the property of small journals run mostly-independently by their editorial boards, are now the property of one of the big journal aggregators, after they bought up all kinds of journals and their archives for relative peanuts.
Dead authors of decades-old basic research originally published their papers where they thought they would be spread widely (and often out-of-band distributed preprints or xeroxes to each-other; it’s not like the authors were excessively worried about copyright or decades-later royalties). But now, through factors entirely out of their control and irrelevant to their original preferences, those papers are being locked away in a paywalled archive to support an aggregator’s rent-seeking.
Century-long copyright is bad enough for books and songs and films. It’s absolutely disastrous for science.
Even today, scientists aren’t publishing in big journals for the “publicity” per se. They’re doing it because they need to for hiring/tenure.
Elsevier and their ilk have basically managed to insert themselves in the middle of a big game everyone was playing where previously money was only an incidental part of the game so nobody paid much attention or considered what would happen if actors with malicious intent got involved. Once firmly in place, they started demanding ransom, and it caught everyone else off guard. Essentially everyone else (libraries, researchers, government grant agencies, universities, the general public) would be happier if they were gone, but for the moment no one knows how to get rid of them.
The tactics employed by the journal aggregators, such as forcing universities to buy bundles of thousands of irrelevant journals they don’t care about in order to have access to the handful with useful stuff in them, are shameless thuggery.