Maybe I'm misunderstanding the abstract here, but doesn't this also suggest that lonely individuals more readily reach their own conclusions about common ideas and concepts? I can't get away from the thought that all this confirms is that groups tend to converge their thinking and speech through regular contact, and that different social groups (including groups of one) will diverge in thinking over time.

well, it didn't say their descriptions were WRONG, just different. The insinuation is that highly social people were informed by what their peers think and say, so their descriptions will likely mirror other socially connected people. The lonely people would just have to come up with it on their own.

I know there's a tendency to dismiss groupthink as negative and wrong and bad, and for huge amounts of people that's true, but for small social groups it's often a sign that you've all become familiar with each other, experienced the same things and are just similar in general, and in terms of selecting for safety, these are all markers of who you will likely feel safest with.

there's a hypothesis that singing and instrument usage like drums came about as a way for a community to show cohesiveness and immediately find out who strangers are. By the time you've learned their songs you're not a stranger anymore, but if you can't sing or talk like they do, you're very likely a stranger to be wary of. Makes a lot of societal evolutionary sense.