From last night’s “Hannity,” here’s the man himself responding to that NYT piece that set conservative media (and grassroots fundraisers) on fire. He’s eager to note how many tea-party candidates he’s supported in the past, both through American Crossroads and his own checkbook, but there’s still no answer to the key question: If the goal is to nominate the most conservative candidate who’s electable, how do you determine who’s “electable”? If you’d have asked me whether Rubio was electable in Florida in 2010, I could have given you an argument either way. He’s electable because he’s charismatic, has state legislative experience, might draw Latino votes from Democrats, and isn’t prone to moronic damaging soundbites. He’s not electable because he’s running as a conservative in a state won by Obama in 2008 against a popular moderate Republican governor in the primary. Which way would the Conservative Project Victory have come down on that? Per Steve Kornacki, which way would they have come down in Rand Paul’s race against Trey Grayson in Kentucky in 2010? Kentucky’s a red state so a tea-party insurgent stands a decent chance of winning the general if nominated, but Paul had “bad optics” baggage for the party in his criticism of the Civil Rights Act. How do we distinguish the disqualifying soundbites, like Akin’s rape comment, from the non-disqualifiers, like Paul’s take on the CRA? Will the answer perchance depend on how likely a candidate is, if elected, to make trouble for the GOP establishment? If you’re more of a libertarian than a “big government conservative,” it’s awfully hard to trust Karl Rove to separate the wheat from the chaff in primaries.

But then, we’re assuming that Rove’s main goal, and American Crossroads’s goal more broadly, is to boost establishment candidates. Is it? Or is this reorientation towards electability more about protecting their viability with rich contributors after a disastrous election year?

But there are plenty of Republican donors who are furious at Crossroads for wasting their money and aren’t going to be fooled by Rove’s rebranding strategy — or his promises that he will get better results the next time around.

I talked to one Republican operative in Washington who put it this way: “These guys took millions of dollars from big donors last year and lined their pockets. The new money will benefit all the same staff, pollsters, admen and vendors. It’s throwing good money after bad.”

Some donors will walk away, but not all. And by backing the most “electable” candidate in every primary, CPV now has a prefab defense to future losses: They can’t be accused of mismanaging their contributors’ money because they’re betting on the horse with the best odds of winning in each race. It’s like a hedge fund switching to a more risk-averse investment strategy after major losses, even though a lot of Republicans who went bust last year were establishment favorites who weren’t so risky on paper. “Electability” is really just Crossroads’s way of reassuring its funders that next time will be different — with the punchline being that they’ll likely be pressured into backing a few tea-party longshots anyway, just to keep grassroots conservatives from making the CPV endorsement a badge of contempt in the movement that candidates grow reluctant to embrace.

Exit question: Isn’t all of this just a variation on the old debate of whether it’s better to back strong conservatives in every race and risk ending up with a very principled congressional minority or better to support RINOs strategically in some states where conservatives are less likely to win and maybe achieve a squishier majority?