which says, "If a senior scientist says something is possible, he's almost always right. If he says something is impossible, he's almost always wrong."

As for retroactively classifying documents, that isn't new. Last week, I watched a History Channel program on the development of stealth technology (I'm pretty sure it was a "Modern Marvels" episode). The program described how the initial paper stealth study (Project Harvey) was unclassified. It was only after Lockheed proved just how much you could reduce a radar signiture that the project was super classified (SCI and black) to the level of the Manhattan Project. I'm pretty sure I remember reading a little story in one of my aviation magazines in the mid 1970s about how the legendary engineer Kelly Johnson was called out of retirement to work on a "low observables" project. It was only years later that I ever saw that term mentioned again.

Without seeing the doctor's criticism document, I can only speculate on why it was classified. If it revealed a system weakness that could be exploited by a potential enemy, it would be legitimate to classify it. By way of comparison, suppose someone wrote a paper that pointed out a weakness in a weapons system like ECM. If the enemy exploited the weakness, our planes that depended on that ECM system would be vulnerable to being shot down.

On the other hand, sometimes papers get classified to cover someone's butt. It's possible the paper he wrote falls into this category as well.