Indeed it would. Making robots "see" in 3D is some pretty hardcore stuff. One step closer to having robots understand what is happening and making educated guesses. I got into a pretty interesting conversation with a colleague of mine who did his Bsc Computer Science at Tukkies. He was telling me about a lecturer who explained the general thinking of a computer. If you are looking at a computer and a computer operated being with the ability to recognize objects is staring back, he might recognize a human. But, if you were to hide your hand behind your back or a wall (or anything like that), the computer will not register that it is your hand behind the wall. For all that it knows, your hand can be a melon or a box. That was quite interesting, especially after i read an article a while back when Kinect was still in developmet, that MS has about 3TB of human movement and body positional data. Could they slowly be solving this problem? It can be quite an interesting debate.