No, you are completely out of the ballpark. This simply cannot be done with a normal camera and requires very advanced hardware. Ignore the camera in Natal, it is only being used so you can see yourself. None of the tracking uses the camera at all, it is all done with the 3D depth sensor. If you hold an oject in front of a normal camera, it can't tell you how large or how far the object it. Natal is capable to tell how you far the object, how big it is, the shape, etc. It basically creates a 3D model of the object and detects its movements.
then you need at least 2 Natals to take advantage of the tech, front and back to at least capture the 3D movement more accurately across the board?