As a general rule, I only buy computer books that are going to be applicable for a long time. For instance, Knuth is always going to be important unless there is a fundamental change in the mathmatics of how computers work--maybe not even then. OTOH, a book on all the latest CPAN modules will be obsolete before they start printing it. I wouldn't bother buying it.
IMHO, the Camel is right on the edge of the rule above. It will last at least until Perl6 comes out (and even then, much of the information in it will still be good). I mirror your concerns about the index--fortuantly, the index in the Camel is very good. My dirty little secret is that I've read very few of the various perlfaqs--most of the information I need from them is in the Camel already (though there are always some faqs that didn't get into the print version, although these are usually on very specific subjects, like compiling perl on VMS).
I wanted to explore how Perl's closures can be manipulated, and ended up creating an object system by accident.
Note: All code is untested, unless otherwise stated