I'm not 100% certain regarding the timeline, but it's starting to seem odd how every article about Oculus seems to marginalize everyone but Palmer and Carmack.
[1] http://interactive.usc.edu/2012/06/12/mxr-lab-members-and-al...
Reality of hardware was totally different. M$ wasnt ready, Project Natal was couple of years away. They scrambled in panic and bought Primesense. Kinect was basically repackaged Primesense dev kit.
Kinect was supposed to be answer to Wii. Problem is there is a huge difference in speed between tracking 4 points of light versus complex DSP on a point cloud. Kinect had a 100-500ms delay between something happening and skeletal tracking on the screen. Kinect One only manages to cut that in ~half.
You simply CANT do responsive input with latencies in the hundreds of milliseconds. This is why Kinect has failed to deliver what $1 billion of M$ PR promised four years ago.
Kinect is great for puzzle games, art projects, slow complex input.
and this might downvote me into oblivion, but is using "M$" still a thing? I don't see how that helps a discussion.
Not very helpful when you want to find something written in the article by Ctrl+F!