I feel like citing raw # of lead in Florida rather than by % to portray the dynamics of the race is misguided.
Just look at the Siena/NYT poll.
https://int.nyt.com/data/documenttools/flpa-0930-crosstabs/16c21b7ab34ed4d1/full.pdf
Vote intention (D/R/I)
Mail (42/24/38)
Early voting in person (34/23/25)
Voting on election day (18/48/35)
In other words, Democrats are supposed to have a lead in both mail and early voting in person. The fact that Republicans are outpacing in early voting so far might mean that either greater % of Republicans than estimated are demonizing mail voting, or that Democrats are over-worried about covid and sticking to mail voting.*
The latter seems unlikely to me, at least enough to the point that it skews the picture of the race. Not a doomer, but I'm not sure that Democrats are doing as well as people are suggesting.
Edited*
There are several problems here:
1) the sample sizes of these crosstabs are gonna be very small in a poll of 700-something people
2) people often lie or change their mind about voting method! there’s no reason not to believe that a significant number of Rs switched from e-day to in-person early voting!
We are also ignoring the NPAs.
