The accuracy of an article depends on the effort (or ability) for contributors to check facts more than its "density of interest". In other words, experts find this easier than amateurs, and for most topics I think many would rather see no article than a crap one. Also worth a conjecture that the interest in writing cannot possibly always be equal to the interest in reading for all topics. If true this is perhaps the ugliest aspect of Wikipedia.

> Programming docs in open source often have errors. Do the volunteers not deserve the "privilege" of offering their time for free?

This is a non sequitur and I don't think I can answer it. What I will say is that the code review process for a pull request usually includes making sure docs are updated accordingly. Much of that documentation might even be automatically generated from structured comments and validated by passing tests such as in API docs. Humans still of course review after that which mistakes even less likely especially since everyone involved is writing about their own work.