Many other commenters apparently take the view that TAOCP isn't much help to a modern, professional programmer.
That's probably true, in the sense that it deals largely with the kind of functionality modern programmer should be calling from a library, not implementing by hand. But I think it's hugely enriching to read these books. I read 1-3 back in the 70s, and was particularly inspired by Sorting and Searching, and especially the chapter on random numbers. RNGs have intrigued me ever since.
I find Knuth's writing style very pleasant. He's funny. So is it worth the time and effort? Not if "worth it" means "makes you a better Javascript programmer". But it's worth it in the sense that reading books about the history and geography of a country is "worth it" before visiting; your appreciation of the place is deepened, possibly transformatively. It could even be that you would never have visited had you not read those books.
I found TAOCP easy to read, much more so than most textbooks. I don't think I tackled many of the exercises; certainly none of the challenging ones. So for me, the "effort" aspect of "is it worth the time and effort" was quite low, so the ROI was quite high. If you find the books hard to read, then don't bother (and perhaps you aren't really interested in low-level algorithms).