I'm not saying that its not a problem, but that it is perhaps less common of a problem than you suggest, and more readily dealt with by existing recruitment and training mechanisms than you presume.
Actually, my essay is based largely on reports issued by the CIA's Center for the Study of Intelligence. I relied on experts who have studied the problem for years. That's not to say that they're infallible, however I don't think that their findings should be dismissed without proof which counters their conclusions.

The same applies to Freeman's research findings. If the brain does create its own internal models, then certain things need to be done to break those models down. The fact remains that each person cannot know what he or she doesn't know, regardless of good intentions or re-doubled effort.