It depends on what you want to do. Here's a simple example: if you want to write games with Unity then basic maths will see you just fine, but if you wanted to write your own competitor to Unity, then you'd need a higher level of maths. Same applies to OGRE, and Bullet, and... the list goes on.

FlexyCore tried the same thing a few years ago with iSpectrum: http://www.flexycore.com/ispectrum-overview.html .They even made the source available:http://www.in-the-box.org/ .

Also sounds like the stuff from eyeSight. I used this a few years ago (on Symbian): http://www.eyesight-tech.com/technology/

My not-so-old house has toilets with a "continental shell" and a tap to do the flushing, So after the Kenyans get their newfangled MicroToilets, I'd like one of their old ones to replace my shitty (pun intended) Western-style pot.

Quite a few years ago I wrote a J2ME Bluetooth racing game (along with an artist friend). We secured a popular license and publisher for it, and the previews and reviews were good (in real printed magazines!). We then sat back waiting for the mountains of cash to be delivered to our doors. It didn't happen. The game didn't get the exposure we were expecting, for whatever reason.

Not to be deterred, we took an improved version of the game engine (since the license was tied to the publisher) and developed something new. Our previous lesson learned (don't tie yourself into stupid deals) we found a new publisher. A small one, but one who assured us premium placement on a popular brand of handsets. The testing period dragged out, much frustration ensued, but the game finally launched. Again, to good reviews (not as good as before, since this time the game was starting to show its age).

It didn't really sell. Never mind, we said, we can take the it elsewhere. So we took it to one of the bigger publishers, who would give us less of a cut but a lot more exposure. The game by this time was no longer cutting edge. In 2004, when it was originally created, the renderer was impressive. Fast forward to 2007 and it looks shabby next to the other hi-end racing titles.

Anyway, reviews were still okay and the game sold in decent numbers. That smaller cut, when going through multiple aggregators doesn't amount to much per unit, but the rise in sales make up for it. Or they would have done if we'd managed to get any of it from the distributor, who a year after the deal was signed went into receivership without ever paying a penny.

Developing games anonymously for the big guys made money. Pushing our own stuff never did.

For the curious, you can grab the game's source code here