Thanks for the reply - your ideas for third generation work that follows from your current research are excellent, and really get at the point that there are multiple different potential levels of intervention that can grow out of second generation work - i.e. how much do you address the overlying context, vs. target the pathways by which they act. As you note, even when acting on the more proximal pathways though it is important to maintain awareness of the broader context and distal influences.
And I agree that the place-based nature of the barbershop intervention was key, especially for communities that may have less trust in the medical system. The social communication piece (i.e. getting health information from trusted community sources) further addresses this issue.