That having been said, the Chicago Manual of Style is in need of a revision. Specifically, it does not cover issues that arise when self-typesetting books or journal articles, a common practice among mathematicians and scientists in the age of LaTeX.
As with any reference of this type, it will take the reader a little time to become accustomed to the order. A first-time user will swear at it, but after repeated use, the user becomes more familiar with the how and why of this work.
Things that at first I found frustrating I now realize could NOT have been handled in a better or more efficient way. There is often no obvious place to put exceptions or obscure rules, and the editors pick a likely location. For example, suppose that while editing, I encounter a situation which doesn't quite fit a standard rule. At first, I think that this exception obviously belongs in Location A in CHICAGO, and wonder why the editors did not put it there. However, a month later, I may encounter a similar exception, but believe now that it obviously belongs in Location B in CHICAGO, and wonder why the editors did not put it there. Later, I realize that I have now thought that the same exception belonged in two different locations -- obviously, the editors can't just keep putting the same exceptions in every possible tangential location. As I gained familiarity with the book, I came to understand why certain exceptions or certain obscure rules were placed where they were -- and I came to agree that they were generally placed in the best location.
That said, there are still a few things I haven't found, but those generally involve simultaneous applications of multiple rules. Each rule is covered, but sometimes, it is unclear how multiple rules intersect.
I am entirely unwilling to trade it my CHICAGO for AP, MLA, Turabian, Strunk & White, or any other style manual.