Patrick Healy writes in today’s New York Times (emphasis added),

In recent weeks, Clinton advisers have been challenging Mr. Obama’s electability in a general election, and her victories in Ohio and Pennsylvania are perhaps her best evidence yet to argue that she is better suited to build a coalition across income, education and racial lines in closely contested states.

But the Pennsylvania exit polls, conducted by Edison/Mitofsky for five television networks and The Associated Press, underscore a point that political analysts made on Wednesday: that state primary results do not necessarily translate into general election victories.

“I think it differs state to state, and I think either Democrat will have a good chance of appealing to many Democrats who didn’t vote for them the first time,” said Peter Hart, a Democratic pollster not affiliated with either campaign. “Take Michigan. It has a Democratic governor, two Democratic senators, and many Democratic congressmen, so it’s probably going to be a pretty good state for the Democrats in a recession year.”

Mr. Hart, as well as Obama advisers, also say that Mr. Obama appears better poised than Mrs. Clinton to pick up states that Democrats struggle to carry, or rarely do, in a general election, like Colorado, Iowa, Missouri and Virginia, all of which he carried in the primaries. Obama advisers say their polling indicates he is more popular with independents, and far less divisive than Mrs. Clinton, in those states.

“Hillary goes deeper and stronger in the Democratic base than Obama, but her challenge is that she doesn’t go as wide,” Mr. Hart said. “Obama goes much further reaching into the independent and Republican vote, and has a greater chance of creating a new electoral map for the Democrats.”

Indeed, if Mr. Obama does become the first African-American nominee of a major party, the electoral landscape of the South could be transformed with the likelihood of strong turnout of black voters in Republican-leaning states like Georgia and Louisiana, which Mr. Obama carried this winter. (Mrs. Clinton has also argued that, given the Clinton roots, she could put at least Arkansas in play in the fall.)

Josh Marshall concurs:

As Patrick Healy explains, it is simply a fallacy to claim that winning a state’s Democratic primary means you’re more likely to win that state in the general election or that your opponent can’t win it. …

… And it’s really not a big mystery that the argument doesn’t hold up because it wasn’t devised or conceived as an electoral argument. It’s a political argument — one that only really came into operation at the point at which the Clinton campaign realized that it was far enough behind that it’s path to the nomination required making the argument to superdelegates that she’s elected and Obama is not.

In a nutshell, Senator Clinton does better in states that are either bound to go Dem in November no matter who the nominee is or Republican no matter who the nominee is. But Obama does better in states that could go either way.