Maybe Barack Obama's highly effective campaign presages a highly effective presidency, but I think Peter Beinart is forcing the point a bit here:
Presidents tend to govern the way they campaigned. Jimmy Carter ran as a moralistic outsider in 1976, and he governed that way as well, refusing to compromise with a Washington establishment that he distrusted (and that distrusted him). Ronald Reagan's campaign looked harsh on paper but warm and fuzzy on TV, as did his presidency. The 1992 Clinton campaign was like the Clinton administration: brilliant and chaotic, with a penchant for near-death experiences. And the 2000 Bush campaign presaged the Bush presidency: disciplined, hierarchical, loyal and ruthless.
Except that to establish this supposed "pattern," Beinart is comparing oranges to apples to pears: Carter gets judged on his persona, both on the campaign trail and in office; Reagan gets judged half on persona, half on policy; Clinton and Bush, meanwhile, get judged on their approach to management and organization. To see how weak this sort of argument is, consider that one need only rearrange the qualities that are being judged to reach precisely the opposite conclusion - that Presidents' campaigns are a poor guide to how they'll govern. After all, Reagan's '80 campaign was badly mismanaged - he fired nearly all his senior staff after losing the Iowa caucuses, remember - yet once in office he was able to run circles around the Congressional Democrats, and match the canny Gorbachev at brinksmanship. George H.W. Bush won the Presidency by playing the hard-nosed partisan on the campaign trail, but he ended up alienating GOP true believers by cutting deals with Democrats from the White House. Bill Clinton ran as a centrist New Democrat in 1992 but tried to govern as a liberal (gays in the military, Hillarycare), before the '94 debacle forced him back to the middle. And George W. Bush ran two campaigns that were notable for their ruthless competence - yet competence of any sort has been strikingly absent from his administration's governance.
This attempt to establish a pattern is no more plausible than Beinart's, obviously - but it's no less plausible either.