There are a couple things we can do:

• decline to review for the big journals. why give them free labor? Do academic service in other ways.
• if you're organizing a workshop or conference, put the papers online for free. If you're just participating and not organizing, then suggest they put the papers online for free. Here's an example: https://aclanthology.org/ If that's too time-consuming, use: https://arxiv.org/

100% ppl need stop thinking big changes can be made "by individuals", this kind of stuff needs regulation and state alternatives made by popular pressure or is impossible to break as an average worker dealing with in the private sector.

Funding agencies have huge power here; demanding that research be published in OA journals is perhaps a good start (with limits on $ spent publishing, perhaps).

I'm sympathetic, but to a limit.

There are a lot of academics out there with a good amount of clout and who are relatively safe. I don't think I've heard of anything remotely worthy on these topics from any researcher with clout, publicly at least. Even privately (I used to be in academia), my feeling was most don't even know how to think and talk about it, in large part because I don't think they do think and talk about it all.

And that's because most academics are frankly shit at thinking and engaging on collective and systematic issues. Many just do not want to, and instead want to embrace the whole "I live and work in an ideal white tower disconnected from society because what I do is bigger than society". Many get their dopamine kicks from the publication system and don't think about how that's not a good thing. Seriously, they don't deserve as much sympathy as you might think ... academia can be a surprisingly childish place. That the publication system came to be at all is proof of that frankly, where they were all duped by someone feeding them ego-dopamine hits. It's honestly kinda sad.

more like the only way to float, not just move up. good luck getting grants without papers in these scum of the Earth publishers

Publishing comes with IP laws and copyright. For example, open access articles should be easy to upload without concern. "Private" articles being republished somewhere without license is "piracy", and ResearchGate did get in trouble for it. It's complicated. https://www.chemistryworld.com/news/publishers-settle-copyright-infringement-lawsuit-with-researchgate/4018095.article

Pre-prints are a different story.

you're risking copyright nastygrams, but people still do it, and even upload preprints and full articles to scihub, because fuck that and it's maybe free citations

The problems are wider than that. Besides, relying "individuals just doing the right thing and going a little further to do so" is, IMO, a trap. Fix the system instead. The little thing everyone can do is think about the system and realise it needs fixing.