You consider those close margins to be leads? Have you ever followed a presidential election? Predicting whose going to win more than two years from now is like predicting when you'll take a dump two years from now.
Those margins indicate that the Republicans would probably lose North Carolina in the next election unless something changes the dynamics of the political scene. The burden of proof is on you to show that something will happen that makes these numbers a poor representation of reality. Sure, there is plenty of time for much to change between now and November 2016.
Hillary Clinton has learned much from her 2008 campaign. She will have the Obama campaign on her side -- and I have never seen as competent a campaign apparatus in my lifetime.
If you predict that Hillary Clinton will lose, then you had better explain how the suggested opponents will do better in a campaign.
