Clearly in a community of unconditional cooperators every agent obtains a better payoff than any agent in a community of defectors.
As soon as you’re talking about communities, you’re talking about meta-PD, not PD, and as I’ve explained above, rationalist agents play meta-PD by making sure cooperation is desirable for the individuum as well, so they win. End of story.
As soon as you’re talking about communities, you’re talking about meta-PD, not PD, and as I’ve explained above, rationalist agents play meta-PD by making sure cooperation is desirable for the individuum as well, so they win. End of story.