
The days of coding your own emulator just for the hell of it has passed. That was back when emulation was something new and fresh. Like many things that become popular, everyone wants to give it a shot. Now, people just want an emulator that progresses nicely (not necessarily fast) and has good compatability with some decent "candy" features like netplay. The general consensus definately agrees with the mindset that multiple groups may as well contribute to the same project rather than seperate ones, as compared to the now-unpopular mindset that competition is good. Many coders agree with this line of thought, and I, personally, also agree with it.
This isn't factoring in the situational specifics of this particular question, even. It is rendered more infeasable by the fact that ZSNES' core is in assembly, and while that wouldn't be too hard to port (both the PC and the XBox have the same processor core, the x86), it wouldn't be a walk in the park, either. With the speed that XSNES9X goes, getting a minor tweak from being coded in assembly versus C/C++ wouldn't even be noticable. In short, it's not worth the time that would be invested in it.