I don't know if there's really such a thing as "peaking too early" with respect to your actual chances of winning the caucuses. There may be if your nearest opponents are going to hit you with negative ads, but that doesn't seem to be happening here (at least not yet). So not sure this will really qualify as "peaking too early". Being ahead in the Iowa polls at this stage is presumably better than being behind in terms of your likelihood of winning the caucuses.
But the second consideration is the expectations setting, and its impact on the momentum that an Iowa win will generate. On that dimension, I think "peaking too early" could be a worry. If you take the lead early enough that you become the expected winner, then you risk not getting nearly the same polling bounce from the victory that you would if the victory was unexpected.
As a rule of thumb, that seems about right to me. But in this particular case, the bump should be big regardless of how obvious the win is. Because even if it's obvious to everyone weeks away, Cruz winning Iowa will turn Trump from a bragging prognosticator into a proven loser, the impact of which will be huge. Cruz will see a doubled surge: from boosted media coverage as a result of his defeat of Trump, and quite likely in supporters, as some of Trump's fans abandon him for a stronger candidate. (And that's assuming Trump fans even get around to voting in the primaries in the first place....)
Poll
