I took the survey.

I was within a factor of 2 on the Europe question, which is pretty good, I think.

As a general rule I "cooperate" on prisoner's dilemmas where the prize is of a trivial size, regardless of my opinion about the incentives and people involved. An interesting experiment might be to take people familiar with the prisoner's dilemma, flip the "cooperate" and "defect" incentives, and see if it makes a difference.