This argument from Ramesh Ponnuru is not very persuasive:

The conventional wisdom about the Republican primaries has repeatedly and rapidly changed. Originally Jeb Bush was said to be the favorite; then Marco Rubio; then we heard that “it’s going to be a Cruz-Rubio race”; then a Cruz-Trump one. The currently rising theory is that Trump will prevail.

Given the poor track record [bold mine-DL], maybe it’s time to let primary voters in a few states weigh in before making confident predictions, or endorsement decisions based on them.

It’s true that many political observers jumped to very early conclusions about which candidates are the most competitive and most likely to be the nominee, but unlike the pro-Bush and pro-Rubio spin over the last year there is some reason to think that a Trump-Cruz race or a Trump-dominated primary season is in the offing. The first three examples Ponnuru gives are all results of grossly overestimating the strength of the Bush and Rubio campaigns, so if anything there should be even greater wariness now of making that same mistake.

Trump and Cruz are the two candidates with consistently higher support in national and state polling, and this holds true for states that vote in March as well as in February. This was true before they started fighting with each other, and it has remained true over the last few weeks. Pundits desperately keep trying to make a Rubio surge happen, but all that this has done is to make him the target of other “establishment” candidates. In recent weeks, he has actually been going backwards because of the negative ads his rivals have put out against him.

If the conventional wisdom about the race has “repeatedly and rapidly changed,” that is because many pundits have tried to find excuses to talk up various weak candidates when they had very little popular backing and before they had done anything to deserve the praise. The state of the race itself has been remarkably stable for weeks, and Trump’s lead in most places has been steady for months and has increased over time. If there is only now a sudden realization that it will be Trump-Cruz contest, that is because party leaders have been too complacent about both Trump and Cruz and failed to do anything to prevent that scenario. Ponnuru thinks they are choosing poorly because they are letting their anger against Cruz blind them. That may be right, but they are also reacting out of haste because they never thought things would reach this point where they would have to choose between two of their least-favorite options.

To some extent, the party leaders that are now more sympathetic to Trump have been misled by the Rubio boosters that have continually promised that he is the “real” front-runner or most likely nominee. That was wrong, and it is practically impossible to pretend otherwise any longer. Party leaders didn’t try to coordinate opposition against either Trump or Cruz because they thought it would be unnecessary. Now that they are being forced to choose between someone they don’t trust and someone they positively hate, they are taking the path of least resistance and choosing the former on the assumption that he’s more likely to win the nomination anyway.

Siding against Cruz at this point is foolish for them only if these party leaders have profound ideological disagreements with Trump, but they are assuming they don’t have any because they believe Trump is conning his voters. That also happens to be Cruz’s argument against Trump, and so these party leaders are ending up more on Trump’s side in part because they think Cruz is right about Trump’s phoniness. Maybe that vindicates Cruz’s indictment of these party leaders, but the open secret is that they don’t care about this. If they have to lose the election (and I suspect many of them are resigned to losing at this point), they may as well do it with someone whose impact on the party over the longer term will probably be minimal. They are betting that Trump is that someone, and on that point they are probably right.