I played Skyrim with Kinect yesterday. It was a blast!
The best part was swapping between my archery/sword/dagger loadouts including switching gloves. It makes the game SO much easier. Saying the names of the dragon shouts made things SO much easier, being able to 'detect life', throw voice and sprint saved time and stopped me from constantly pausing to access the favourites menu. Also, saying 'quick map, Whiterun' was handy too.
As far as down sides go there were few. I enjoyed the Kinect functionality with Mass Effect but it felt a bit too sensitive, if my wife was having a phone conversation the next room over it would often trigger powers. With Skyrim however the Kinect seems tuned a lot better, I hardly had any interference except the one time the wife said 'sorry' which caused my Dragonborn to cast a frost spell. After playing a couple of hours I think I only needed to repeat myself about 5 times when the game didn't understand me so that's pretty good in terms of accuracy. I found that saying the shouts in the dragon language was almost impossible, the game couldn't differentiate 'fas' and 'fus' so it was much easier saying 'unrelenting force' and such.
Anyway, it was a great experience and it made the game a lot more enjoyable. Having bought the Kinect for the kids I didn't expect to use it much myself but I found the 'you are the controller' slogan gaining some traction. I didn't feel like a controller as such but I did feel like I was an extension of my controller and I felt like my voice was far more effective than any hotkey or quick select menu. It seems so intuitive bypassing half a dozen buttons in favour of saying a word. It makes me excited about the future of controllers and our bodies being peripherals or extensions, in particular with voice commands. I'm not talking just with Kinect but the concept as a whole. I'm not interested in using my body as the sole control input as I agree with the sentiments expressed on this show that a controller is suitable because it allows a great range of complex actions to be mapped to minimal button inputs. But if you reverse that philosophy and look at how complex games require lengthy input and user interface to achieve simple results like swapping a weapon, doesn't it make sense to use our body's infinite repitiore of simple gestures to bypass encumbering user interface?