I think there are a couple of big problems with glass:
1. It doesn't do AR substantively, which is the breakthrough feature, instead its basically just a semi-transparent floating screen
2. Its got a photo taking camera on it*
3. Interacting with it is cumbersome/clunky
To do 1 effectively, it needs to be full field of view or at least the bulk of the FOV, which is a major issue with lenses that close to the eye and why I think they acquired Magic Leap.
2 is a fairly simple fix but would need a whole new re-branding and google may have burned this bridge already.
3 is really a tough thing to figure out but I think could be done with CV gestures that aren't physical touch based, but rather view orientation based. So for example if you look down, you immediately get your apps displayed and you can use your hand to select/type in a similar form function as people do now with phones.
*I make the distinction about photo taking because if the camera was only used for CV/tracking on device and had no "photo taking" feature, people would be less worried about it. And yes I am aware that images are still taken, its the distinction in what is happening by default that the public in general cares about.