Let me try an angle here by way of hypothesis:
The first thing to note about the advocacy for the Cruz vote in Iowa is that it was largely (though not exclusively) based on work by Steve Deace. No problem with that. Central to the conclusion were a few things
1: The polls which were predicting a high turnout, were wrong.
2: Trumps expressed preference support as revealed by polling wouldn't turn out.
3: Smaller turnout favoured the established caucus goers and the existing networks of the politically active.
4: A large turnout would be indicative of people who hadn't participated previously, and by extension therefore more likely Trumpsters.
So this is the question you'd like to ask. If you were somehow armed with the information before the vote that Iowa's Republicans would generate a turnout higher than any previous one recorded (180,000) would you still have recommended Cruz at 6/4? I'd suggest that you probably wouldn't. Yet this is what happened, and Trump was defeated.
So how did this happen? and can we extrapolate anything from it going into New Hampshire?
Only about 10% of the qualified population caucus. To give it a UK context, think police commissioner elections. Did the fabled silent majority find their voice? Well it's not like these folk were lost to the political process. Iowans do vote in greater in numbers than 10% after all. Have those who normally vote but don't bother with the aggro of a caucus come out to oppose an anti christ candidate in greater numbers than new participants have been drawn into the process? If this process repeats in other states, then Trump, sitting on a series of commanding leads heavily odds on, might suddenly become vulnerable right across the board?
So what of New Hampshire?
In the first case the group who are perhaps most capable of organising a mass turn out by way edict are the Godists. They'll exist in New Hampshire too, but not on the same scale as they do in Iowa. Having said that, turnout is the key. You don't need a big mass movement to affect disengaged constituencies on small turnouts. If you can mobilise a minority you can still achieve a lot (I have no idea what sort of turnout an NH primary generates?)
Perhaps more pertinent though is the spectre of Pat Buchannan. Standing on a ticket of anti immigration (with a bit of Holocust thrown in for good measure) he managed 31% of the vote in 1992, and won the New Hampshire primary in 1996. Trumps more charismatic than Buchannan, but then so is typhoid. Also the mood swing post 9/11 you'd assume would make an electorate with whom the sentiment had twice struck a chord in the 1990's even more receptive to the message. The final thing that would worry me is how the opposition might mobilise behind a stop Trump candidate. Kasich, Christie and Bush will all likely be hitting close to double figures. There ahs to be a limit to the size of the pool you could draw from. You suspect that those looking for a moderate alternative to mobilise behind won't find one for now, so is Cruz still over the odds at 7/1 if we see something of a repeat performance on another high turnout?. The evidence from Iowa if we're prepared to trust it, is that high turnout hurts Trump and is more indicative of opposition to him than support gathering.
My gut feeling is that Iowa won't repeat in New Hampshire. I suspect there's enough latent sentiment behind Trump already there, and no natural point to rally against him either. It's going to require three quite capable candidates who've been working the patch to collapse for Rubio to make significant inroads. And is a Texan evangelical really going to get traction there? South Carolina however, now there's somewhere which might be more vulnerable. Sadly there's no mature market for it yet, but if Rubio is going to breakthrough, it will need to come soon before the doubts are raised. Can he really run with form figures of 3, 3, 2, 2
