As someone who wasn't around when the GP2X scene was at its height, maybe I can offer a slightly different perspective. I'll be honest, I bought my GP2X (from eBay) primarily as an emulation platform, and for that it's outstanding. It has more emulators available, for a wider range of machines, than any of the other open handhelds -- including the BBC Micro, which I don't think any of the newer open consoles support at all.
So, I probably spend 80-90% of my GP2X time in emulators. I *do* play original games as well, but they're up against very tough opposition. For example, Head Over Heels and Beneath a Steel Sky are both freely available and easily run on the GP2X under emulation. Unless a game is so good it really blows me away -- Animatch is an example of a game that *is* that good -- then I tend to play it a few times and then forget about it, and go back to my emulated games.
One of the reasons I started my blog was that I was hoping to give a little more visibility to GP2X game (and non-game) programs, but even after several months of reasonably regular posting, I'd had virtually no feedback -- ironically, the same problem the programmers had! -- except for the few people who commented on the forum thread in here, and most of those were people who'd been around the scene for years in any case and knew much more about the games than I did.
It's a bit of a vicious circle, I think. I look at the archive and am frustrated by the *huge* number of programs that could have been fantastic if only they'd been finished. (Even in the emulation category -- the PSX emulator, for one.) There are a lot that have messages like "Don't worry, I intend to keep updating this!" -- but not a hint of any *actual* update for years. In at least some cases, I'm sure that the programmers lost interest because they weren't getting the feedback... but perhaps they weren't getting the feedback because the game hadn't been finished!
The sum total of my GP2X programming experience consists of an SDLBasic program to draw Lorenz attractors, and even that took me a while, so I don't underestimate the difficulty in porting games from other platforms. Take the GP2X version of Jazz Jackrabbit -- it's certainly not what I'd call a finished, polished game, but there's no way I could have done any better. The problem is that there is no longer the critical mass of interest to keep things ticking over -- and, unlike with the Spectrum, the userbase wasn't huge in the first place.
If there had been a million GP2X units sold, then who knows what would have appeared? I'd guess there'd be quite a few nice OpenGL games, a decent X-based graphical desktop, lots more ports of things like Midnight Commander (that worked!) and lots more besides. After all, consider that the much, much less powerful Psion 5MX can (with some messing around) run Debian Sarge. Would we have ended up with complete distros being adapted for the GP2X? We'll never know, sadly.
Okay, I've rambled off the topic here (as I tend to do) but I suppose my point is that doing those things would have felt worthwhile to many more of the talented programmers and designers who've owned a GP2X over the years, if only there had been a sizeable user base *that lasted* to support it. The Spectrum was commercially viable for a full decade, and had a huge user base. The GP2X wasn't, and didn't. There's your difference.
