I'm using "research" broadly. People can privately research and publish under whatever arrangement they wish. Academics too. For those that sell their copyright, that would be valued based on projections of what the copyright buyer expects to sell. For those whose research was funded by the copyright buyer prior to writing because they didn't want to bear the risk or lacked the funds, that's going to be less, but everything stems from how the books are expected to sell.

I think there are better funding models than the precarity of copyright-dependent publishing.

Yes they do. As a kid our (admittedly not particularly lavish) family summer holiday was paid for each year by the modest royalties my father, an academic, received from a couple of academic textbooks he had authored.

Right, that was worded too universally, but this is what I mean: royalties here are supplementary to a salary, not a wage in themselves. The research was funded before publication.