What's So Stunning About Clinton's (Narrow) Win in New Hampshire?

On the NewsHour this evening, Jim Lehrer called Hillary Clinton's win yesterday in New Hampshire a "stunning comeback." Several dozen other news sources echoed those words, including the Chicago Tribune, the BBC, the Financial Times, and The Washington Post, calling Clinton's win "one of the most stunning comeback victories in modern American politics." MTV News called it "a shocking win."

Stunning? Shocking?! I was mildly surprised, but not stunned or shocked. Clinton won, but not by that much. Is the MSM overreacting or just trying to make the New Hampshire results seem more newsworthy?

My friend, Peter, texted me at about 9:30pm EST last night:
WTF - Clinton closed an 11 point gap? Bad polls? Her loss in Iowa energized her base? Sold her soul to Satan, again?
The 13 point swing -- from down 11 to win by 2 -- did seem big, but polling data has been predictably unreliable in recent years. Remember those exit polls in 2004 that predicted Kerry would beat Bush? (We'll never know for sure whether Bush won fairly in 2004 or 2000, but that's another story.)

A friend emailed me the following this morning:
Why in the WORLD are all the pundits saying Hillary "won the New Hampshire Primary" and is "back on top." She and Obama tied, both receiving 19 delegates from New Hampshire, and Obama still has the overall lead in number of delegates. Where are the pundits/journalists/talking heads getting their numbers?
I was wondering the same thing myself. So I took a closer look at the numbers.

The difference between the two top finishers was just 7,479 votes. If anyone had told me a month ago that Obama would lose to Clinton by just three points in New Hampshire, I would have scoffed.

Clinton, Obama and Edwards all collected delegates, so I wouldn't rule any of them out based on these numbers. If anything, I would argue that it's a three-way race, but that's about all.

So why were the polls wrong? There is the conventional wisdom, courtesy of the New York Times. I would add that poll respondents were reacting more to Iowa than to the candidates, and pundits assumed that the voters in Iowa and New Hampshire were more similar than they really are.

How will voters in Nevada and South Carolina respond to the results in Iowa and New Hampshire? I'm now left wondering, can we expect more "stunning" results in the coming weeks?