Which is just a tautology.
Hotline Miami - 2 core devs + 2 others; Risk of Rain - 2 core devs + others; Gunpoint - 1 dev (Tom Francis FTW, yay!) Valdis Story - Dunno, but website says "we" and mentions a few different people in blogs; Terraria - 2 core devs + others; Vlambeer - 2 core devs + others.
It is very difficult, almost impossible for a single developer to make a competitive game now. Not actually impossible, but the dev needs to be almost preternaturally productive. It is much more common to find programmer + artist + others teams, as it became after the transition to 16-bit. Very common to have programmer + artist for pre-alpha prototyping, contracting additional programmers + additional artists + sound guys (as for several of the games you cite) + platform conversion teams + music, etc.
> pixel art using GameMaker, for example.
Pixel art can have high production values. Pixel art was still the main requirement for 16-bit games, well after the demise of the bedroom coder.
If you think that, say, Risk of Rain doesn't have many person years of production input, you're deluding yourself. Try making something equivalent yourself this year.
For the record, I don't think pixel art has to be particularly high quality for the game to be successful. Right now I'm playing an alpha of Vagante and having a ton of fun. It's looking like it's gonna be a hit. The graphics are muddy and kind of NES-looking with lots of cheap rotations — and that's OK.
> Try making something equivalent yourself this year.
I'm actually planning to do just that. So I guess we'll have to see!
Right, them's the 'production values'. They aren't comfortable because the quality bar is pretty high, even on pixel art, high enough that you really need an artist to do it, not a lone coder.
I think the situation isn't quite as bleak as he says in the OP anyway. The age of the lone coder is gone, I think. The few exceptions are exceptions. But the age of the prototype-in-a-game-jam-get-crowd-funded-and-build-a-product-with-a-small-team isn't over, though the bubble has definitely burst from when every kickstartered game seemed to reach its goal. So while I think he's right that the market has eliminated the lone-coder approach to game development, the ability to bring a good small game to market is as good as it has ever been.
As you said further up, it is PC that is the heartbeat of that.
Good luck on your game.
Minecraft succeeded not because it was one out of a hundred block-based games that got lucky, but because it had a perfect mix of simplicity (slick interface and controls), immediacy (play in your browser), wonder (explore forever in any direction), and persistance. I guess you could point to Infiniminer before it, but if you've ever played it, you'll know that it was slow and clunky in a 90's sim-game kind of way.