my problem with natal is that it is been only a vague demonstration so far...
the motion recognition system, can it be overthrown by other movements in the room? what is the range of the recognition? and the voice as well, it would be incredible if it just sorts through all the words you may throw out to recognize specific ones or will it end up mixing commands up?
and the other problem is that it may look 1:1 now, but i remember when seeing the wii for the first time, many thought it was 1:1 or close to it after just glance but after dealing with it for a while, the problems only appeared more and more-- hopefully natal was not overly polished and buttered up for this show because if it works, then it easily destroys nintendo...