I'm pretty sure they discovered/made things they could leverage with regards to integration with apps/smartphones and voice interaction. They did a lot of work with interactions involving very pared down and minimal visual information. Also, it doesn't have to stop with a smartwatch. What if they had an in-car HUD as well as a background service running on a smartphone that could leverage the HUD, the watch, or any enabled and linked display device? For example, what if he whole system could figure out I was in the living room, then enable a Glass-like interaction through the media PC+remote, with optional voice? (You'd have to activate the voice, of course.)
Re: camera -- I can see lots of potential for far more discrete devices and telepresence devices, especially in corporate and government environments.
2. Not sure how displaying notifications and having a camera with some basic voice recognition is that big of an advantage right now for a smartwatch. Maybe I'm missing something big, but it doesn't seem like Glass did anything that could translate to a smartwatch that smartwatches don't already do better than Glass. Also, they already have Android Wear.