An Economic Geography of the United States: From Commutes to Megaregions (2016)

Then I think you would probably enjoy Colin Woodard’s “American Nations - A History of the Eleven Rival Regional Cultures of North America”

Measures of cultural flows or how much cultural affinity people feel?

Cultural flows may be more feasible if one were to look at vectors for inventions, dishes, or ideas (i.e. distance travelled over time) as a proxy for connection.

How close people feel to their neighbours I would say has more to do with specific communities than actually how connected people are. For one thing, there would be the assumption that people are aware of what other communities are like or doing in order to have any set stereotypes or sense of affinity.

Honestly, I have dreamed and imagined how that could happen for years now. Of course, you’d have to start somewhere just like this. I think Facebook data with image recognition and and text analysis would be my first choice, largely related to Facebook “Groups” and menber join dates. Overwhelmingly imperfect but would be quite interesting.