The smart watch proliferation is interesting. It's a new kind of device, for which a new theory of UI design will be needed. Where a phone can be used with two thumbs at once and can show a fair number of items on screen at a time, a smart watch, because it's worn on one wrist, can only be used one-handed, and can probably, in a pinch, show up to four touch targets at once. It's hard to write software in those constraints, but it's possible.
Is it good enough for most interactions for most people most of the time? Probably not. I think smart watch owners won't ever be able to use it as their primary computing device, unless their needs are very limited. If you just make calls and get reminders, maybe that would work. If you send lots of texts, though, I don't think it's going to work very well without an alternative keyboard. This might finally be the format where we have to do away with QWERTY and go to dictation for most tasks. So maybe you could get away with a watch and headphones. Maybe. I think you'll still have a computer, the way most people don't just own a phone or a tablet. A watch is a terrible primary screen.
I'm prepared to see that change in time, though. Maybe, as smart watches get people used to the idea of wearing their computers on their wrists, we'll start seeing sci-fi-like forearm computers with bigger screens, or maybe some other form of interaction will take over.
Smart watches can do a lot, and I'm glad they exist, but we can't think of them as little phones. I think they need a different mode of thought for them to work properly.
Mokalus of Borg
PS - By my count, this makes at least five basic form factors to design UIs for.
PPS - That's watch, phone, tablet (touch), desktop/laptop (mouse & keyboard), television (remote).