The same author may approach different publishers depending on their intended audience. e.g. Daron Acemoglu published "Introduction to Modern Economic Growth" with Princeton but "Why Nations Fail" with Crown (a division of Penguin RandomHouse).
* SICP: MIT Press
* Modern Compiler Implementation in ML: Cambridge University Press
* Creating Symmetry: Princeton University Press (https://press.princeton.edu/books/hardcover/9780691161730/cr...)
It really depends on the nature of the book. If it's a first-year undergrad text that is likely to be purchased by thousands of students, a publisher like Addison-Wesley will be interested. If it's solidly mass market applied tech, like "Master Javascript in 32 Seconds" or something, someone like O'Reilly or Manning will probably take it. If it's more niche and technical it's likely to be a University press.
University presses do publish some absolute bangers. Creating Symmetry, for example, is an absolutely beautiful book combining maths and art. MIT has a great series on game design. Cambridge is very strong in machine learning and maths.
I do find that my philosophy and critical theory books are the ones that more heavily come from university presses. In general the humanities tend to have stronger presence in the university presses (largely because books, not papers, are the publication format of choice for those areas).
It’s a lot easier than many folks suppose to reach the bottom of what the Internet has on a lot of topics—people toss around phrases like “all the world’s knowledge” and, oh boy, is that ever not true—and IME the next step is very often a few books from university presses.
But when you need tenure, and you need to get published, they're the 1000x more willing to publish you than one of the major popular presses.