So a textbook is useful to the extent that it helps its reader pass a class or a test. Whereas a (non-fiction) book is useful to the extent that its author manages to convey his or her reasons for setting out to write the book in the first place; for conveying what they found interesting or remarkable, not (as determined by the bevy of textbook publishers) what's gonna be on a test. I agree. I think that I took a slightly incredulous tack earlier because I imagine a chemistry, or even a psychology, textbook to be prima facie and less controversially useful, more than a textbook on history could be, although it is history that this entire post is concerned with. Thank you for the insight. And like I said, I am an abiding fan of non-fiction. Already on page 87 of Destiny Disrupted.
Textbooks are didactic. Non-fiction is autodidactic. This is by design: "what math do I need in order to understand physics?" is an externally-guided search. It also holds true for liberal arts: "What songs should I listen to in order to understand the influence of music on American culture?" is not a question you can solve yourself without a whole bunch of research and exploration. However, "what are the cool parts of physics?" becomes an autodidactic search because "cool" is a controversial qualifier. Likewise, "I want to know more about Bruce Springsteen" is not a quest that should be informed by committee. This is an autodidactic discussion. "I'm interested in exploring my historical blind spots" is not something that should be answered with textbooks. It's like suggesting someone learn more about the events that led to September 11 by reading this thing.