Subscription models make publishers insist on controlling access to research they didn't perform, write up, or fund. They act like a midwives who insist on keeping (or hiding, or performing surgery on) other folks' babies.
Taxpayers pony up $28 billion annually for NIH to fund medical research, resulting in 60,000 annual published studies. First beneficiaries of that knowledge aren't doctors or patients, but journals that are prohibitively expensive for many.
"Academic literature should be freely available: developing countries need access; part time ... thinkers ... journalists and the public can benefit; ... you’ve already paid for much of this stuff with your taxes ... important new ideas from humanity"