Qualified people will have major incentives to produce contributions and have them accepted and be accredited for that work
Is this because of the prestige (including career benefits) of contributing to "the" textbook for that subject?
I agree wholeheartedly on the benefits of open-source textbooks, just like open-source software. However, if top-quality contributions aren't being commonly donated, I'd have no trouble having the books cost something. That is, while there would be freedom to create and distribute derivative works, both these and the originals wouldn't have the freedom to read without paying. The problem would be enforcing this, as broke students would tend to free-ride if there were no consequences. This is unlike business users of open commercial software, where there's more pressure to obey the licence.