Michael Barone’s first mistake is to use numbers for Gallup to try to make a credible and convincing argument. Gallup lost all of their credibility in 2012 and they can’t begin to get it back until they get good results in an actual set of election predictions.
Barone’s second mistake is to think that a regional look at the president’s approval numbers can or will translate to a Democratic nominee. Hillary Clinton, for example, simply sells better in many states than the president. That was true in 2008 and it will be true in 2016. If some other candidate emerges as the Democratic nominee, they are likely to have a different profile of support, too. I’m thinking of Brian Schweitzer of Montana, for example.
The same thing is true for any potential Republican nominee. Bobby Jindal and Ted Cruz are going to have limited appeal in the Mid-Atlantic or Pacific Northwest, but an unscathed Chris Christie might have done well in those regions. Mike Huckabee will never win a state in New England, but Jeb Bush might have a fighting chance in Maine or New Hampshire.
How the people feel about President Obama will matter in 2016, but it will a minor factor compared to the impact of the actual candidates.