I'm more curious how the display works. I had Google Goggles. The display was nearly impossible to see and gave me a headache going crosseyed trying. If the hardware can overlay well, and be easy to see, someone will write the software for it eventually. If it can't, then no amount of software will work.
Really the glasses itself should be dumb. Put all the smarts in a smartphone app with a plugin architecture. Have it voice controlled where you say a keyword then parse your next sentence for a command. Send the video to the phone for processing via BT, and let the app do whatever it needs to do, and individual plugins send overlays up. That's enough to get a lot of useful stuff out of it.