This a comic about... Well, just read it. It's quick.
Until something noteworthy happens, this is probably the last time we'll need to talk about Kinect as a general topic. I sort of can't resist is as a thought experiment. As I suggested initially, different developers often have radically different notions regarding usability - and there doesn't appear to be a stable conceptual platform that defines the "Kinect" experience. Even beyond that, different developers approach even the fundamentals according to their own needs.
The DS is essentially a dense knot of sensing capabilities, so broad that I couldn't make sense of it at first. Between its d-pad, its eight buttons, its microphone, its touch screen, and its connectivity suite, capabilities just pulse there in the dark awaiting a master. Many games allow you to play with or without buttons, as you like. The Kinect is the closest analogue to that. It's just a big ol' pile of technology, and it's not going to make sense in absolute terms until the right person arrives and pulls it from the slab of legendary stone in some whispering, primeval forest.
What is generally considered the most noteworthy game for it, Dance Central, doesn't even leverage the Kinect's much vaunted ability to gin up a wireframe of your body. Ubisoft's Motion Sports doesn't appear to use that either. The choice in both cases was to use its capacity as a depth camera to cut players out of their environments, tracking their silhouettes as input. They do this (presumably) to minimize latency, and it works for this purpose. Dance Central gets out of the player's way almost completely, attempting to create an experience that is as unmediated by technology as possible. Part of the reason this game is brought up so often, and not just by me, is because it actually works. I don't mean works from a technology perspective, that it functions. I mean that it vindicates some portion of a subset of the theory that games without controllers can exist, and Microsoft didn't even make it. It's such early days for their intimidating, onyx surveillor, and I suspect that in a year's time they will discover that their every assumption about "best practices" for the machine were wrong.