Ross Douthat sticks to his guns and outlines how Trump could be beaten:

Think about it this way: It now looks very likely that Cruz will beat Trump in Iowa, at which point Carson’s campaign will be pretty much finished, and Mike Huckabee and Rick Santorum will disappear — and with Cruz suddenly ascendant, he’s likely to pick up their supporters, pushing him up to Trump-like levels in the national polls.

Once that happens, would you really bet on Trump to beat Cruz in many of the evangelical-heavy Southern states that vote in late February or early March? In Cruz’s own state of Texas? That seems unlikely.

There are a few holes in this scenario. First, Trump leads in several Southern states that have been polled so far, so it’s entirely plausible that Trump could prevail in many of these primaries after winning New Hampshire. Polling in the fall even had Trump slightly ahead in Texas. So, yes, I would bet that Trump wins many of these primaries. It is still likely to happen even after a Cruz win in Iowa. Trump is ahead in South Carolina, and he is far enough ahead right now that putting all of Carson’s support behind Cruz doesn’t change that.

Besides, I wouldn’t assume that all Carson supporters flock to Cruz. Most of those that want to do so have probably already done that, and many Carson supporters are probably more likely to favor the other “non-politician” over the senator. Some Santorum and Huckabee supporters have reportedly been considering an an “anyone-but-Cruz” strategy on the grounds that many social conservatives (correctly!) see Cruz as untrustworthy, so it is unclear why the few supporters these candidates have would automatically go to Cruz anyway. Cruz can and probably will do very well in several Southern states, but can he win anywhere else? My assumption is that Cruz is primarily a factional candidate of “very conservative” voters, and that limits his opportunities to win.

The March 1 Super Tuesday includes Massachusetts (where Trump leads by a wide margin), Oklahoma (where Trump and Carson were 1 & 2 in the fall), and Georgia (where Trump has been up by double digits). He’s also in decent shape in Ohio, Missouri, and Florida, all of which vote in mid-March.

Douthat also overestimates the ability of an “establishment” candidate to thwart Trump in later primaries:

At best, two of those three men will be left as viable candidates after New Hampshire, and the Republican establishment (such as it is) will simply have to pick one to rally behind by the end of March. The complexion of the race will be different if Christie somehow shoulders past Rubio — but either one of them should be capable of stopping Trump from sweeping the blue and purple states he needs.

It’s not clear why either of these candidates “should be capable” of doing this. If they haven’t won anything before the end of March, it is not obvious that any of the “establishment” candidates will have a rationale for continuing to run. It will be too late by then to stop him in blue and purple states, because the “establishment” candidates that might theoretically be able to do it would probably all be out of money or so far behind in the delegate count as to have no chance.

Not only does Trump tend to have support drawn pretty evenly from different parts of the Republican coalition, but his support does not appear to be limited to just one or two regions. Trump picks up as much or more support from moderate and “somewhat” conservative voters as he does from “very conservative” Republicans, and that gives him a better chance in more places to come out ahead. Blue-state Republican primaries seem designed to thwart someone like Cruz, and Trump seems well-positioned to do very well in them. And if Cruz isn’t going to be the one to beat Trump (and he isn’t), it’s hard to see who would be.