And Arena was made by a tiny team of less than half a dozen designers/programmers and half a dozen artists. They only made games for the Xbox. I know nothing about programming, but I seem to recall plenty of people talking about how the PS3 cell processor was difficult/different to program for, so there was no simple process to just recompile and pop out a Playstation version of something. They didn't make PC games either - they were Xbox programmers. Expecting it to be multiplatform or that it should have been concerned with capturing a larger audience misunderstands the scope of the team and what their capabilities were.
Hmm, that reminded me of something else that I overlooked earlier because I actually managed to forget what a big issue it was: namely, the engine. For a developer today, making a game for the Xbox and then porting it to the PlayStation is just a non-issue - whether they're using Unity or Unreal, the engine basically handles all the difficult bits, and all you need to do is basically check if there aren't any unexpected graphical issues And even so, there are large companies that specialise in porting games, because this process, though less complex today, is time-consuming and resource-draining. It used to be much worse though. Much worse - because most developers at that time really had to develop their own engine first. There was no commercial engine that really fit the bill. Torque was still around in 2006, but it was barely viable. Unity was first released around 2005, but it wasn't really viable for bigger developers before roughly 2010. Unreal was available, but back then, the license fee was a six-figure cost that added a very significant overhead to a small project, and didn't really guarantee success, because Unreal was an engine for FPP/TPP shooters, not the amazing universal tool it is today. I mean, I remember our own engine issues at work precisely in that 2006 timeframe, and basically the conclusion was that unless we're making an FPS, we're best off continuing to develop our in-house engine, because the options for air combat are just so limited.
And yes - as you said, converting to the PlayStation came with its special, unique idiosyncrasies. As late as 2012, when we were finishing up Dogfight 1942, the process of preparing the game for the PS3 was a nightmare, even though we were using a commercial engine that theoretically supported the PS3 (keyword: theoretically). We had two programmers out of a team of about ten spend months and months trying to clear up all the kinks that emerged in the process... and from what I can see in Arena's credits, Gaia had something like three full-time programmers altogether, and they were probably working with their own engine.
Come to think of it... I mentioned earlier that Arena would have been developed on a PC, and therefore it would probably be easy with the source code to rebuild the game specifically for the PC. But that may have been me thinking about the project as though it were a present-day game. Back in 2006, the devs may well have taken all kinds of shortcuts to keep the budget down, and those shortcuts may have meant the game can't run on PCs, or can't run on them very well. It may well be that today, it would actually be easier to sort of reverse engineer Arena and remake it on Unreal or Unity using just the graphical assets. But then we come back to Arena being somewhat less than popular - who'd have the time and dedication for a project like that?