It occurs to me to point out that Who Voted for Trump and What Voters Caused Trump to Win are two different questions.
The former question is one of straightforward broad demographics, which of course show you a snapshot at a moment in time, and can’t show trend.
To answer the latter question, you have to take the subset of Trump voters who voted for Obama in 2012, and run a demographic analysis on THAT, plus how that subgroup rated the issues.
From what I have seen, the Dems do not seem interesting in analyzing that slide of the electorate. At their peril, because if that demographic continues to trend GOP……
They appear to be intoxicated with Trump sitting at about 40% on the RCP approval scale, and willing to forget that (a) Obama spent a not-insignificant amount of time during his presidency at that same level, and (b) Obama spent most of 2012 at or around 45% before winning a second term.