My fundamental objection to these devices is that there isn't enough of a UI change to fit the form factor they're on.
One of the reasons Windows Mobile was kinda lousy was because they tried to shrink a desktop OS down onto a phone (and now they're making the same weird mistake by scaling a mobile OS up to the desktop). Fundamentally, these are different things, so they need different ways to interact.
These watches seem similar. A mobile OS that's been massaged a bit, but realistically makes more sense on your phone than on your wrist. Watches were (are) great for the amount of information they deliver in the time they deliver it. That is, a small snippet of info, delivered incredibly quickly. Your interaction with it is minimal. For a few things, like meeting notifications, maybe these smart watches make sense. But for emails or texts, they fall down a bit in my mind because you need more information than is readily absorbed at a glance.
One of my other problems with them is that except for notifications, they don't really make you any faster. Assume that you get an email, and your watch lets you know. From the demo slides I've seen, the watches let you see some information, delete the email, open the email on your phone, or start composing a reply--also on your phone.
For the last two operations, you have to pull the phone out anyway; I'm not sure what you get by starting to open them while you're still staring at your wrist. Deleting the email from your watch is potentially useful, but that's a bit of a stretch; I don't think I've ever seen an email that it's urgent to delete.
So, yeah. The problem isn't with any of the things you mentioned, the problem is that I think we're still trying to figure out exactly what the watch could possibly be used for, and basing the operation on the operation of a completely different device with a wholly different interaction model and parameters.