nickeubank
commented
3 years ago To be clear, visualizing the spatial distribution of polling places is a great way to make sure your data is coherent any specified things well. it just doesn't do a very good job of telling you about where polling places are more or less available given the density of people. 
Generally, without some way to normalize by population, all "count" maps just become maps of the US population, so much so there's an xkcd for it:
https://xkcd.com/1138/
To map meaningfully, you'll need to do something like calculate "polling places per person" for counties, then color counties by polling places per person.