The region is only "known to be involved in processing social information". Saying that "the results suggest social closeness is more important than shared beliefs when evaluating others" is a long long shot just by observing that the region gets more active with friends. Our brains could simply be more active because with our friends we have more memories that our brains needs to look up, process and relate with whatever input we're getting from them. I didn't have the chance too read the original research papers but saying that "The results suggest social closeness is more important than shared beliefs when evaluating others" is too vague. I also wonder whether this phenomenon also happens with friendly people.