If you watched the video (or just the part I suggested) I'm curious what you thought in general about it.
Wow. Smart guy, and that may be part of the problem. No manager directing them. The scientist have been allowed to get lost in their self absorption
Seriously, overall feeling that explains a lot of what doesn't work well, is that the scientist were allowed to do whatever they deemed best, without practical direction.
I was troubled with the opening, basically it doesn't need to be easily discoverable (WTF?) just easily learnable. Wrong. Somewhere along the way a Manager of technical people should have stepped in and said make it easy to discover and easily learnable. It is interesting that after he made this statement he qualified it by saying "as far as 'I'm' concerned."
Take the second lady where they show the demonstration of her selecting tiles. I agree, tile selection works well. But she said "now that I know how to do it ..." OK, so why waste everyone's time trying to discover how Microsoft's software engineers decided this should work? Especially considering they designed it under the mantra that easy discovery is not important ... Just tell me or show me then, and don't waste my time. I really got the feeling a lot of intellectual masturbating occurs at Microsoft Research in the touch/gesture department.
I did think they were on the mark with a lot of their biomechanical assertions. Yet, he said the first lady chose to put the slate on a pillow because she felt that was the best way to interact with it. WTF? She did that because in that position, that's the only way to avoid a back and shoulder ache.
Then the speed bump thing, that's good stuff. But this is critical, it causes the tile to move slower than the finger. This in and of itself can make many gestures unnatural and difficult to learn especially if there are any hardware performance issues which I believe is why I naturally began to tap and hold the tile for a brief moment before repositioning.
I basically stopped halfway through. Clearly they are great engineers, but I think their perception is skewed by a disconnect with the real world. They are beginning to interpret human behavior the way they want to see it to fit their technical theories instead of shaping their technical theories to fit human behavior.
At the beginning of the video he states "teaching users of our apps gestures that work well with our systems." I think the Modern UI will evolve much better if that changes to "creating gestures, systems, and apps that work well with our users."
Lastly, he opened by stating these are ideas, "the start of something." I agree wholeheartedly, feels like unpolished ideas waaaaay too often.