I don't fancy getting into the weeds on this one. I used to do statistics, but no more. I'll wait for other people to look at it and report. After seeing the graph by Simplicius I recalled my 'Eureka' : ) moment, thinking it does not smell right, but I could be wrong : /. Larry J, although an analyst, perhaps fell into the same trap as myself.

Not knowing much about US elections despite reading that book by Greg Pallast, and assuming your figures are correct, the figure of -6 million voters i.e. not voting seems significant. It's possible one might want to examine that to see if there has been voter 'suppression' .. perhaps .. and compare etc. I'll report if I see something more promising.