There is a hidden assumption in the analysis, and that is that Democratic (remaining) in-person voting on election day in 2020 will be similar to their in-person voting in 2016.
I believe that in a normal year that would be a reasonable assumption, but this is not a normal year.
This is the year of the coronavirus, a disease that has the amazing ability to terrify Democrats but not concern Republicans that much.
That is why I am predicting the remaining Democratic vote will substantially under-perform what we would normally expect because the Democrats will not want to expose themselves to coronavirus from fellow voters and poll-workers.
Imho that should assure that the President wins _all_ the “swing states” and a few others as well.
We will know soon enough.
Also, at least 1/4 of the attendees at Trump rallies are democrats. Do not assume that a democrat ballot is a vote for Biden.
Move Iowa, North Carolina, and Arizona red on that map and it’s President Trump for 4 more years.