I agree with Steve Kornacki that Sarah Palin would be a potentially disastrous candidate for the Republicans in 2012. People really don't like her, and it's difficult to see how that turns around; moreover, she gives every indication of being a massively subpar candidate, unable or unwilling to learn enough about public policy to avoid a new series of damaging gaffes.
Kornacki's jump from there to Walter Mondale in 1984, however, doesn't really work. Yes, Mondale got clobbered -- but he was clobbered by a popular incumbent president boosted by a strong economy. Under those circumstances, it didn't really matter who the Dems ran; when the electorate is happy with the incumbent, there's not very much the out-party can do about it. In fact, I happen to have a tab open with a paper by Larry Bartels and John Zaller looking at the effects of economic and other variables on presidential election results, which shows not only that Reagan's victory was exactly in line with the economic variables, but also that once all the objective variables are tossed in that Reagan actually slightly underperformed (see figures 3 and 4). Now, one of the variables included in their analysis is ideological extremism, and I don't know how they coded Mondale; presumably, a more moderate candidate would have done somewhat better. However, the magnitude is pretty small; in the text, Baretls and Zaller single out Goldwater, McGovern, and Reagan (in 1980) as the three examples of ideologically extreme candidates, and estimate that such extremism costs about three percentage points. So just on ideological grounds shifting from Mondale to, say, Gary Hart, might have been worth one point or so, maybe.
The rest of it? Mondale was a perfectly fine candidate, in a year in which Democrats really had no chance.