I use mine on occasion. It's particularly useful when targeting architectures that don't have modern intrinsics for special operations.

It's worth noting that some of the solutions aren't optimal for modern devices with conditional moves. In some cases you can get equal or better performance with straightforward code than some of the stuff in the book that goes out of its way to avoid branching control flow.

Ha ha, this falls firmly into the "I wish I needed to reference this more" category for me as well.

Same. But sometimes I like skimming it just as a distraction. There's something so satisfying about seeing these tricks I'm far too stupid to come up with on my own. A couple of times, I've had some breakthroughs when stuck on an unrelated problem, while reading this book, because the "psychedelic hackery" sort of leads me down an alley I wouldn't otherwise have thought of, due to some superficial similarity.

When I was doing low-level optimisation for games (lots of SIMD), I was using mine quite frequently.

Now that I'm doing more application development it's sitting on my bookshelf looking neglected.