Right now, I'm pretty sure most people don't care about academic publishing or even know what is going on. The scandal in my opinion is that a professor or phd candidate is paid by the state to do research and then the state has to pay again to license the contributions their employees made. Meanwhile the taxpayer who already paid for the research cannot access that research for free.
The main problem is that the academic system is a dog eat dog world with people clinging to their positions or fighting with all dirty tricks in the book to get tenure, reputation or funding. I'm not sure how to fix it.
The way to get to a better world probably leads through blood sweat and motivation by small groups. Run an open access journal, make it the best or highly regarded in your small niche. The more of these exist, the better.
Maybe that's an interesting area for an accelerator to pump some non-profit money into. Hire some people as editors, make the content available for free. Use startup strategies to turn it into a very feisty publisher where academics want to publish. Make sure the stuff in the journal is easy to cite and becomes cited often etc. Use a bowling pin strategy to start with one topic (probably something computer science related) and once that is excellent, branch out.
You can run a better journal for free than what you can get from Elsevier. Throw in forums, video conferencing and shared docs and you can build entire academic structures in the cloud. Run a journal on a subreddit, papers on arxiv, video on meet (recordings posted to youtube), code in github, computation on colab notebooks. WeWork should offer wetlab space. ;)
Ideally we would have a storage commons where data that doesn't fit into the arxiv model of papers. Something like object storage but with some built in structure.