Eric Ham, author of The GOP Civil War, argues in The Hill that Jeb Bush has an advantage over Hillary Clinton in the electoral college:
Clinton has the path of least resistance to the Democratic nomination. Yet a head-to-head battle with Bush could spell doom, as his advantage in the all-important Electoral College is unquestioned.
His evidence for this theory is weak at best. His arguments are that 1) Bush remains popular in Florida; 2) Colorado elected a Republican Senator in 2014; 3) Nevada sent a landslide of Republicans to office in 2014, including popular Latino governor Brian Sandoval; and 4) Republican governor John Kasich remains popular in Ohio, could be a vice-presidential pick, and the GOP is holding their convention in Cleveland.
This is thin gruel indeed. First of all, 2014 results will likely have about as much bearing on what will happen in 2016 as 2010 results did in 2012. There are two very different electorates in the United States today: those who turn out in presidential years, and those who turn out in midterms. It’s not as if Ohioans and Nevadans changed their mind en masse between 2010 and 2012, or between 2012 and 2014. While opinions differ on Hillary Clinton’s strength as a candidate and her ability to motivate the Democratic base, Democrats had the same qualms and concerns about Obama’s re-election in 2012. It’s possible that left-leaning voters might be lulled into complacency by an ostensibly moderate candidate as they were in 2000, but a third Bush candidacy will be almost certain to drive Democratic base turnout against him because of his family name alone.
The other problem is that even if you assume Ham’s postulates about the way states will lean, the math still doesn’t work for Bush or the GOP. Chris Ladd’s famous analysis in the wake of the 2014 election bears repeating: there is a gigantic blue firewall at the presidential level that will be impossible for the GOP to overcome absent an implosion by the Democratic nominee or a shocking lack of Democratic turnout.
Let’s look at an optimistic map for the GOP:

Let’s grant Ham’s premise that Bush’s background in Florida will help any more than Gore’s background in Tennessee aided him. Let’s also grant the dubious notion that the Kasich effect and the GOP convention will somehow put the GOP over the top in Ohio. Let’s give the GOP Colorado despite its ever-more-blue demographic shifts, and let’s assume that some combination of Bush’s Spanish speaking and Sandoval’s endorsement somehow pulls Latino voters in Nevada to Bush in spite of the GOP’s rabid freakout over immigration and unwillingness to budge on the issue.
It’s a series of dubious longshots, but let’s give Bush Florida, Colorado, Nevada and Ohio. Let’s also give Bush the swing state of North Carolina (because this isn’t a discussion otherwise), and let’s assume that 2016 isn’t the year that demography overwhelms the GOP in Arizona and Georgia. Let’s also assume (as is likely but not at all certain) that every red-leaning swing state like Missouri and West Virginia falls to Bush.
Even all that still doesn’t get the GOP to more than 268 electoral votes. The Democratic nominee would still win the White House, even under that highly improbable scenario. Bush would have to go beyond that feat to somehow pluck off Iowa, Wisconsin or Virginia just to eke out a narrow win–nor is there any particular reason to believe that will happen. Scott Walker will be dragged through the mud by grueling presidential politics, Virginia continues to solidify its Democratic core in the north, and an older white Democratic nominee is likely to perform significantly better in Iowa than even Barack Obama did in 2008 and 2012.
It’s not impossible for Democrats to lose the 2016 election, of course. Any number of things can happen to make a nominee unelectable. But absent some kind of massive implosion by the Democrats, the Republicans will still need something like a miracle just to keep it close in the electoral college. Jeb Bush certainly doesn’t have an electoral vote advantage in April of 2015, whatever may occur in the future.