I also have a theory that early voting has FURTHER skewed the polling in favor of Ds because the pollsters have failed to reduce their D samples in all states to acct for D EV advantages. In NV and IA this could be significant because of the actual # of D early voters. However, in OH, where the sheer #isn't great, the Ds apparently are using up Election Day voters ("cannibalizing").
An ACCURATE poll would reduce D pools by a small %, THEN ask first, "Have you voted?" If yes, ask party, then ask if they voted in last election, then in the 2006 election, then ask did you previously always early vote or is this something new for you" If the latter, BINGO. you would know that you have a "motivated/likely voter" who this time voted early and then won't be there on Election Day, then do the math.
Watching FNS and no one there or any others ask the question of how many D’s are voting R this year. It is like the elephant in the room. I’m not a D and never have been but that has GOT to be a factor this year. Just like it was in 1980.