This is highly suspect. One need only look at the source material for Trafalgar Group's results in North Carolina to see why these polls are questionable. 31% of respondants were from the Raleigh DMA. 27% were from the Charlotte DMA.
But the Charlotte DMA is larger than the Raleigh DMA, so why would Raleigh be overrepresented compared to Charlotte in their polling? And, of course, white people are overrepresented in this poll, at 68.5%, whereas non-Hispanic whites make up only about 53% of the population of North Carolina.
And what's with all the respondants who don't even live in North Carolina? You're really polling people in Norfolk, VA and Spartanburg, SC about the presidential election in NC? 12% of respondants were from out of state?
That's just one of the cited polls. What do you want to bet they all have problems like this? And that's before we get to the fact the Democracy Institute is an offshoot of the Cato Institute, which is itself a Koch propaganda instrument.