Two predictions which seem to be undersold

But obvious.

1) The health insurance mandate on employers will be avoided by having companies pay the fine and put their employees on publicly funded options. The only thing stopping them is reputation scoring. Tossing thousands of people out of their benefits at the earliest opportunity does seem a little harsh to be done all at once. But it will probably happen. The math and profit margins are just too good to be ignored. I would say this is a good thing, since it will mean the end of the employer coverage mandate and the tax preferential status we've accorded that for decades for no apparent reason. Except it will probably mean several hundred billion dollars per decade in additional public spending instead of saving us billions of dollars on health care as it should have (by shifting more responsibility to private consumers for modest health care demands).

2) COIN doesn't work when the local population does not trust the local government. Or rather the non-local, national government. That our "success" in Afghanistan is, for bizarre reasons, predicated almost entirely upon. We should really learn if we're going to invade other countries to prop up tottering democratic looking systems that we'd better make damn sure to not pick the people ourselves. I suppose we could also learn not to invade other countries to prop up democratic looking systems. But that might be asking too much all at once.