The WSJ's Gerry Seib has an excellent survey of the good and bad scenarios. Here's my vote for most plausible good-case:
The forces of repression win within Iran, but international disdain compounds, deepening world resolve to stop Iran's nuclear program and its sponsorship of extremists. In other words, Iran doesn't change, but the rest of the world does.
And his most plausible bad-case, in my view:
There is some legitimate recount or runoff, but Iran emerges with Mr. Ahmadinejad nominally in charge anyway. He emerges beleaguered, tense and defensive, knowing he sits atop a society with deep internal divides and knowing the whole world knows as well. His control is in constant doubt. What's the classic resort of such embattled leaders? Distract attention from internal problems with foreign mischief, and use a military buildup (in this case, a nuclear one) to create a kind of legitimacy that's been shown to be missing on the domestic front.
Seib also closes with a critically important point that I haven't seen much elsewhere. If the Obama team is tentative or uncertain about this situation, that's in large part because America lacks good intelligence about Iran. Administration officials lament that, much like in Iraq before the war, we have few spies or even back-channel ties there--and thus the Obama team relies heavily on the media in its effort to divine what the heck is going on.