Can Romney, under the right conditions, break out of the 20s, where he’s been stuck for almost the entire campaign? Sure. Jonathan Bernstein is persuasive on this point: Polls of Republicans show that Romney has low negatives and he performs well in hypothetical match-ups against other Republicans. But what if Romney faces the wrong conditions?

The central mystery of the Republican primary is this: How can Romney’s support be so stable in a primary that’s so volatile? So far, we’ve seen a boom-and-bust cycle take Michele Bachmann, Rick Perry and Herman Cain from candidate to frontrunner to not-gonna-happen. But after each bust, Romney’s support was unchanged. He picked up none of their disappointed supporters. Now he’s trailing Newt Gingrich in the polls. Here, via TPM, is what that’s looked like in Iowa:

Ron Brownstein has come up with perhaps the best explanation for this. The Republican primary, he writes, has “become two races running along parallel but very distinct tracks.” One is the non-tea party primary. There, Romney is winning, and easily. The other is the tea party primary. There, Romney is losing, and big. Here’s the graph, which was published before Gingrich’s surge:

But ultimately, there’s really only one primary, and Romney needs more supporters if he’s going to win. And though Romney does not have a low ceiling, he clearly has a sticky floor. Republicans may not refuse to support him, but there’s strong evidence that a substantial number don’t want to support him. If Romney is inevitable, they’ll come around. But what if, at the wrong moment, Romney is not inevitable?

Last night, Romney sat down for one of the first televised interviews he’s given in this campaign cycle. It was with a fairly friendly audience: Bret Baier of Fox News. And it was such a disaster that this morning, the Democratic National Committee released a video splicing together the reviews — many of them from Fox.

Debates are a format that suits Romney well and his competitors poorly. So far, Romney hasn’t even stumbled. But eventually, he will stumble. Nobody runs a truly perfect campaign. So imagine Romney loses Iowa, as is very possible. And, under the strain of the loss, he gives a bad interview, or has a testy debate performance, right before New Hampshire. That might be all the excuse a critical number of New Hampshire voters need to coalesce around Gingrich, or perhaps the excuse that some resigned Romney supporters need to jump ship to Huntsman. And so Romney either loses New Hampshire or barely wins. And then he loses South Carolina.

To be sure, Romney could, even under those circumstances, mount a comeback. As Nate Silver points out, there are eight weeks separating the New Hampshire primary from Super Tuesday. In 2008, there were merely three weeks. The 2008 calendar favored a momentum candidate like Gingrich, while the 2012 calendar favors a fundamentals candidate like Romney.

Even so, Romney is having enough trouble adding supporters that he’s clearly vulnerable to a run of bad luck or bad news coming at the wrong time. And thus far, the primary has been so focused on a medium in which he shines — debates —that his flaws in interviews, his vulnerability to ads portraying him as a flip-flopper, and his weaknesses as a retail politician haven’t really been tested. Romney has looked so strong that even a slight stumble could be significant for a media that wants a horserace and voters who clearly want to support another candidate. So is Romney the likely nominee? Sure. Inevitable? No.