Funny story
The guy who developed EyeToy for Sony (which is a POS based on the most trivial motion detection algorithms) also developed the "Z-Cam" which does, essentially, exactly what Kinect does.
And *they* looked at manufacturing costs, and decided - get this - no-one is ever going to pay upwards of $100 for a piece of hardware like that.
Typical Sony fail.
I doubt Kinect 1 was ever going to be higher-res though. It is like you say a cost issue, but ballpark cost estimates can be made on day 1 of a project based on web searches or general knowledge, so the CPU and camera were always going to be that level of quality.