I completely understand your point. Misinterpreting correlation with causation is something that frequently bothers me in writings and discussions. In this case, however, I did find the correlations the author pointed out interesting. I did not interpret this as causal (the author may have made that claim, but I was more interested in the maps).

I still found it really cool to see how much an impact geography and path dependence has on our current state. I may be over generous, but when they show maps like the Alabama one where an ancient sediment deposit is correlated with farm size, racial makeup, and election results I think they are giving an example of how there can be common causes tied to geographies over long periods of time.

You're completely correct that simply overlaying the Austrian Empire's borders on a map of Romania's election results does not prove a causal link. It is much more likely that there is an underlying common cause (geographic feature, ethnic makeup, etc.). Pointing out these correlations is fun to me, as I'm able to speculate on the possible common causes.