Salon.com is running an election strategy piece about what Obama has to do to beat McCain in November. It's written with the assumption that Obama will get the nomination, but it unwittingly serves as a potent demonstration of why the super-delegates would serve their country better by backing Clinton.
To get to 270 delegates, the piece says Obama absolutely, positively has to win the following battleground states (electoral votes in parentheses):
Michigan (17), Minnesota (10), New Jersey (15), Oregon (7), Pennsylvania (21), and Washington (11).
Clinton and Obama will likely have split those states between them, but Clinton's wins were worth 53 electors, versus 28 for Obama. That emphasizes the way Clinton's popularity is more optimally distributed to win a winner-take-all electoral system like the general election. She'll likely lock up nearly twice as many of those votes easily as Obama will, leaving her more time to focus elsewhere.
The article goes on to list the states that aren't "must haves" individually, but among which the nominee will need to compile a number of wins to beat McCain. Of those, the big prizes are Ohio and Florida, with 20 and 27 electors respectively. It's possible to win without either of them, but to do so a candidate needs to collect a lot of smaller states.
The columnist ultimately decides on 4 scenarios for Obama to win the election:
1) Win Ohio.
2) Win Nevada, New Mexico and Colorado
3) Win Nevada, New Mexico, New Hampshire plus one other small battleground.
4) Win two out of three of those "N"s, plus either Colorado or Virginia.
Of those scenarios, most suggest that Clinton would have a better shot, based on primary results.
She won Ohio by double digits. She won Nevada by six points, despite it being a caucus state, which favors Obama, and despite him winning the key union endorsements, which should have given him a big advantage in a caucus, where union members can intimidate fellow members into compliance. Clinton also won New Mexico and New Hampshire (the latter at a time when her political obituary was running in all the papers).
Colorado was a win for Obama -- but should be taken with a grain of salt since it was a closed caucus system, so the general election will certainly look a lot different. As for Virginia (which went to Obama), it is for Obama what Florida is for Clinton -- the ace-in-the-hole big state that will probably break Republican but has a chance of breaking Democrat and changing the whole game.
Obama can definitely win the general election -- but the risks are higher than with Clinton. Among battleground states, Clinton likely would have more electoral votes locked up pretty easily (freeing her to concentrate elsewhere), and have an easier route to more electoral votes in the toss-up states. And let's not forget that the last two elections came down to Florida and Ohio, which both favor Clinton over Obama by double digits. For Clinton, winning Ohio is very likely against McCain, and if she does so, the election is all but won. If she doesn't take Ohio, taking Nevada, New Hampshire, and New Mexico plus one other small state is a pretty realistic outcome.
So, I see two relatively smooth paths to her victory. With Obama, I see Democrats having to bet more heavily on peeling off traditional red states like North Carolina and Colorado... and I see greater risk of McCain peeling off traditional blue states like Pennsylvania. Obama can win it, but the math is a lot tougher for him.
It's a question of popularity distribution. Clinton and Obama are statistically identical in terms of overall popularity, but Clinton's generally less popular in the blue-lock and red-lock states, meaning she's got more optimal distribution of the same amount of popularity in places it might actually matter.
Also remember that there will be no caucuses in November. Obama did a terrific job inflating his delegate totals by winning 13 out of 14 caucuses, where his more "enthused" supporters could buffalo people into backing their candidate. But he lost most of the regular balloting states, including nearly all the big ones. He doesn't get to use his caucus advantage against McCain. It's kind of like betting on the Super Bowl when one of the teams was nearly unbeatable at home, on artificial turf, with the surface favoring its team strengths, but which lost most of its games on grass. If the Super Bowl is played on natural grass, that's cause for concern.
Finally, remember that battleground demographics also favor Clinton. Sure, Obama will handily win the youth vote and the black vote and the liberal vote. But what Democrat doesn't? The real question is what happens when you offer swing voters their second-favorite Democrat, and then offer them the alternative of McCain? Hispanics, older people, centrist Dems, and the white working class all have an unsettling risk of breaking for McCain, so I'd think we'd be in a stronger position if we offered them their first choice nominee, not a consolation prize.