The problem here with Glass is it's not really a Sousveillance device (in this context). It's just going to allow slurping up vast streams of data for the benefit of Google (and secondarily for the wearer). This is but one of the reasons I'm more of a fan of Steve Mann's EyeTap tech (which is wearable, personal, mediated-reality computing) - it allows for the idea of Sousveillance to be enacted by the wearer.
Sounds great for the wearer to choose where they want to upload their data, but everyone else will still be subjected to continuous surveillance. At least currently this is a rare, unacceptable thing to do, reserved for sting operations and the like.
You want EyeTap. I don't want your surveillance. We can't both have what we want.
Sousveillance, not Surveillance. There is a distinction, which is basically one of being anti-Surveillance.
But of course, with Google Glass, the concept of Surveillance is being somewhat muddied and spread across the people that will be using it.
And you should look into how Police forces are being more regularly equipped with on-person Surveillance.. I'd prefer to have my own version of events that can be compared with the "authorized" version (where you can insert your own value of "authorized" if you want), and I'd prefer it to be under my, not Googles control.
Sousveillance only really counts when it inverts the usual power imbalance inherent in authority-led surveillance systems.
The problem is that if wearable systems like Glass or EyeTap become widespread, this will create a new power imbalance between those who use it and those who don't, raising the same issues brought up by mass surveillance systems but on a more personal scale - which rather spoils the whole ethos of sousveillance that makes it so attractive in the first place.
edit: I do keep harping on about it, but his book "Intelligent Image Processing" is worth a read for more info on how the Eyetap works (amongst other topics).
(edit: for clarity)