The NYT seems to think that NC looks like a 6 point win for HRC. http://www.nytimes.com/interactive/2016/upshot/north-carolina-early-vote-tracker.html?_r=2
But they also admit that there prediction is based, in part, on their polling.
Indeed, on Saturday, total early votes (in person and mail in), Republicans were -15% behind 2012, Democrats were -8% behind 2012 and independents up 25%. Today, Democrats are down 11%, Republicans down 9% and indies still up 25%. This trend is going to continue unless the expansion of in person polling sites drastically increases Democratic early voting and not R. And yes, early voting helps Dems here - Obama had like a 300k vote lead going into election day (lost by 3%)
If you assume the 30% increase in indies, which is nearly 50% male (very good mix) and nearly all white (85%) skew Trump, it's possible that rather than a 300k hole on votes on election day, Trump has a 100k hole. But like I said - we'll find out more over the next couple of days as early voting places expand (in 2012 it was expanded the whole time)