The problem is the base vote. I'm not talking about vote flips, just basic turnout. The polls are based on what they think will be turnout, but if the Rats don't vote in those proportions it won't be that close.
From what I can read and ascertain early voting is not going well, and that's the Rats bread and butter. The younger vote, another Rat mainstay, are the hardest to motivate and don't seem like they are for this one. The union Rats have left or are just plain dead.
The Indies don't seem convinced, and their ethnic group constituency is not exactly as solid as most elections. A loss or shift of 10% of those will leave the Rats grasping for sand.
And the problem for the Rats is that at a general 5% response rate, most polls can't predict that kind of thing accurately in the era of few landlines and a solid reliable way to contact a broad sampling of voters. There's just too much murky undercurrent anymore that can't be pulled out for an accurate survey.
And if that margin for Trump gets around 10%, cheating becomes dicey. There are only so many dishonest election officials and so many precincts (most if not all urban) they are over where they can manufacture votes, and so much they can stuff the box to offset the margins in Trump favorable counties. Or for that matter, a narrower than usual Rat margin, which is actually a far bigger danger for the Rats, especially if it's a big pop county.
All that is to say is that I have no trust in the polls, but the rate at which the Rats are spontaneously freaking out, something is afoot. We'll just have to wait and see.