1. If Perry manages to come in third in Iowa, is it possible to spin this as a major 'comeback' even if people like Silver already are predicting him to finish third. And, if Gingrich performs poorly in Iowa and New Hampshire, how quickly until he is out of the race entirely? Or does that even matter?Two parts.
First, predicting spin. Is it "possible" to spin a (sort of) expected 3rd place finish as a major comeback? Sure; there's only a loose relationship between some sort of objective interpretation of a primary result and what spin gets adopted. See, for example, New Hampshire 1992, when Bill Clinton declared himself the "Comeback Kid" for finishing second, just as the polls had predicted. The thing about Clinton was that he was the early favorite in New Hampshire, but then was hit with a variety of major scandals, so that in a sense his survival at all was perhaps a comeback. Or at least that was a plausible spin. Why was that particular interpretation accepted? Partially, if I recall correctly, because the Clinton campaign did a great job of working the media; partially, I would guess, because Democratic party actors certainly strongly preferred Clinton to Paul Tsongas, and generally preferred him to Tom Harkin and Bob Kerrey.
So it's possible that Perry could be the big "winner" after finishing a 3rd in Iowa that the polling already indicated; whether that would happen or not depends on campaign skills, party choices and influence, and media biases.
Second, Newt dropping out. Well, we can't get into politicians' minds; all we can do is look at the incentives. Just as an aside -- that doesn't mean that I think the incentives are all that matters; it's just that it's all we really can assess as outsiders. And when it comes to personal decisions such as run/don't run it really does get personal.
OK, so: should Newt finish 4th or lower in Iowa, he'll have no chance to win the nomination...but then again I never thought he would have a chance, and wouldn't even if he had won in Iowa. Would Newt believe he had a chance? The polling in upcoming states may take a while to adjust...candidates who surged in the polls once may believe they will again, or they may persuade themselves that the nonsense about this cycle being highly unpredictable is true. And that's not to mention my brother's theory that Newt actively doesn't want the nomination.
With normal candidates who are running in order to win the nomination, it's fairly easy to predict this stuff: they run until they believe they are beaten, which usually takes the form of their donors telling them that they won't give any more. But for candidates who are running for other reasons, whether it's issue advocacy or enhancing one's brand in order to sell more books and get higher speaking fees, it's a lot harder to figure out when they have an incentive to quit.