The linked article says something like: Superrationality is not necessary. All you need is to realize that in real life, there are no single-shot PDs, and therefore you should use the optimal strategy for iterated PD, which cooperates in the first move.

That's simply refusing to deal with the original question and answering something different instead.