Yeah, one of the received wisdoms of the console market is that optional accessories don't really work, so you have to ship everything in the base SKU.
The main problem that I see in MS' strategy with Kinect is that it's tailored to work in a very specific environment (in this case, the main living room of a decently sized house), whereas previous consoles could work wherever you had enough room to place the console, a seat, and a TV. I'm pretty sure that MS' market analysts have concluded that the market segment that has room for an Xbone is also the market segment with the most readily available cash to spend, and so we get a product made for that segment of the population.
It's once more a decision that is completely logical from an engineer's POV.
EDIT:
It seems to me that MS has fallen into the programmer's pitfall of "Let's rewrite this thing". It's somewhat similar to what they were convinced of with Windows 8, where they thought that they had found a new formula, a new paradigm for the future, and tried to get a headstart on the implementation before the other players realized that MS was right. MS has built a machine for ubiquitous broadband net access and mature motion capture control schemes, and a market made up of people who have fond memories of playing stuff on their Wiis; unfortunately for them, that segment is currently rather small.
They tried to innovate in a very conservative business during a recession, they made the wrong choices, and now they're paying for them.