Great additions You are right about even classified strategies being political to some extent. The are LESS partisan, however. Not non or bi-partisan, just less partisan. Part of what made the final Bush 41 NSS so good was that it was significantly less partisan then previous ones due to the fact that it was crafted to assist the incoming Clinton Administration. A second reason was that the managing editor was COL Geoff Jones on the NSC staff, one of the brightest Army officers I've ever met with a really good sense of the political.

I would also add to predictions, presumptions, and assumptions the notion of inferences. Inference can often be confused with presumptions and assumptions (as IMO Steve Metz does in his recent book). I would also note that inferences are less than predictions but may be a critical way station.