Xbox is an odd case because it's essentially a Windows PC on an Intel CPU. A good lot of the code can be executed directly, largely saving the trouble of performance issues and compatibility hacks. Yet, nobody seems to care about it (dare I say there is a reason for that? Ha). There is Cxbx, which is an established project, but the progress is
sloooooooow.
PSP is actually much better: there are several playable games and development seems to be
quite active. I'd suggest monitoring PCSP/JPCSP's progress; in a year or two it will likely be mature enough to run most commercial games.
In any case I doubt any of these newer systems are harder to emulate compared to such monstrosity that is Saturn with its 6-something separate processors working in unison. The problem is lack of interest/expertise. 8- and 16-bit systems have been around enough to have been thoroughly documented, so writing a basic emulator for them using the extensive data accumulated to date is not a huge challenge for somebody with coding experience. With stuff like Xbox there's still a lot of experimentation and insight involved to make things less hackish, or even at all working, so development needs proficient coders for that task. And we all know how hard it is to find a proficient coder with required experience
AND required free time
AND required motivation to work on an emulator they likely won't even use themselves.
