It's alla bout the balance. And even if you can't compare pro games with freeware ones for obvious reasons, the same principles go for both.
Whether you're a solo developer making a game in your spare time, or you manage a team of 200 specialists, you have to manage resources - talent, time, energy, coordination efforts. For an indie developer, time is the crucial factor, for a pro team, it's money. Comes down to the same.
Now you need to balance gameplay vs. graphics, and there are various strategies in doing so. Take "X - beyond the frontier" for example. Graphically, it's far from spectacular, the controls are somewhat iffy, missiles are close to useless. But the storyline and depth are rather convincing, making the graphics "good enough". The other extreme is for example a game like "Speed Busters"; the story is close to non-existant (not necessarily a problem for a racing game, but anyway), it's frustratingly hard to beat even the second level, you're constantly knocked out by random obstacles, but it's full of eye-candy. Weather effects and whatnot. It's nice to play for an hours, after that, you've seen it and it gets boring.
Now Frenetic Plus - I like the smooth controls and the consistent style across the whole game. Making the same game in hi-res would very likely require much more effort for both graphics creation and programming, since one would have to render 4x as many pixels (or more). Making the game just as responsive and just as good-looking wouldn't even be worth the effort, I suppose. So for this game, it is probably the best choice.
And there's one more thing; pretty much every commercial game on the market uses some kind of more-or-less-photo-realistic art style, simply because 3D hardware is commonplace now and this is the easiest way to create impressive graphics when you have a large enough team. But there are other art styles that can be are just as nice to look at, and lots of people like them, if only for the nostalgia factor.
