Regarding ctags, it's useful enough, but it doesn't do a great job of things like jumping to the definition of the "flags" part of "p->flags", if you've got lots of structs that have a "flags" in them. Same goes for member functions. It's absolutely no fun trying to find out what something is, and just getting a list of every single thing of that name in your program. Thanks ctags... but I actually know how to use grep. And if I used grep, I'd get a callers list too.
I found ctags workable, but overall disappointing.
Are you referring to this line:
"The fuzzy parser supports C, but is flexible enough to be useful for C++ and Java"
I think you might be misinterpreting the wording there. It's called "c"scope because it was initially written for C and C only. The line above is suggesting that it also has support for C++ and Java (because it has a "fuzzy" parser... whatever that means).
May clang cause people to expect more from their tools.
I've used various versions of emacs over the years for most of my programming (mostly C/C++ with some sprinkling of python and java). I've recently had to use Eclipse for Android dev. I agree that it'd be almost impossible to integrate the entire GUI dev framework of Eclipse Android environment in emacs (not to mention continuous compiles and other java-firic features) and that's the reason I use Eclipse for Android. But "Jump to Source" ? Emacs is a far better and faster experience for me when it comes to trolling around in big, unfamiliar code-bases. The small subtle lags with eclipse (it's not as bad now with my monster new dev machine with 16G of ram) still irritate me. And the Emacs keybindings are almost there but not quite.. which is another big reason I've resisted IDE's. They're just slower for actual editing of code. Always. Without fail.