It might just as well be intended to establish how many people are so entrained on "cooperating is virtuous in PD-like problems" that we choose cooperation-like choices without actually thinking through the consequences.

I wonder, now, what a typical audience would select if offered a standard PD problem with the labels swapped.