As a maths teacher who also have written a textbook, I have seen a drastic increase in errors in books compared to 5-10 years ago. I don't know why, maybe the scheduler is tighter or proofreading is now too expensive.
They get "typeset" by SPI in Chennai by a non maths speaker, non-native English speaker and come back with errors in both the equations and words and author queries like "equation missing but we follow tex".
One professional mathematican I know in a suitably obscure area of number theory has started her own journal to get around it all. I'm in both the physics and medicine camps and have to put up with this bullshit. It's awful.
I would like to see a peer review/editing community for wikibooks btw, that actually would lead to better quality of books. distill.pub eg IMHO is great.