However advanced Kinect's speech recognition is, Microsoft has a general Speech SDK out there, which means that with a bit of programming, we can get speech recognition tools working with any microphone.
Granted, we won't be able to build advanced functionality since we don't have access to the game's code, but using programs like autoHotKey, we could simulate key presses and combinations of them. You could for example call "TIMEOUT" into your headset, and have the tool translate that into the press of the Back button.
Is it simpler and more safe to use the Back button instead? Sure. Would it be a fun thing to try out to see if possible? GOD YES.
I'll start working on it come October hopefully, but I want to see whether people would be interested in something like this, what functionality they'd like to see in it (remember, you have to suggest something that can be done with a key/button combination), and whether they'd be willing to donate towards me getting a Kinect to try this out with.


So, what do you think?