Paywalled, so I haven't checked my question
3) Do people from the University of Cambridge underestimate self-segregation due to US population mobility?
(FWIW, local politics in my country does anecdotally skew rural/right urban/left, but we fortunately have many other different divisions whose axes are uncorrelated with that particular one. I firmly believe having more than two major parties helps immensely on this front.)
Incidentally, my ultra-low-dimensional model after 2016 was "how far away does a US voter live from the largest city in their watershed?" It remains to be seen if this simplistic metric predicts anything useful in 2020.