If we take the 93% certain to vote for Republicans and 80% certain to vote for Democrats and guesstimate that means exactly 93% of the 26% of the electorate that are registered Republicans will vote, and exactly 80% of the 29% who are Democrats, this would mean about 4% more Republicans than Democrats would vote (which would translate very roughly to about a 1% to 1.5% advantage in total popular vote--depending on ratio of Republicans and Democrats to others voting).
These are rough assumptions, but the picture is very good for Trump if one considers he seems to consistently lead among independents. But I have also seen numbers suggesting he might not do as well among Republicans as Hillary does among Democrats...although I think the difference is not as strong as the difference in independents.
My guess is that if the election were somehow today, Trump would win. But I suspect on November 8 he will win by a wider margin than he would today. My suspicion (which I have had for months) is that Hillary would poll mostly ahead until the debates when she would collapse and Trump would win in a landslide. As far as I can tell my suspicions have played out pretty well so far...
I think party turnout will be GOP 85% to, max, 90%; Dem will be 87% or below.