I'd really love to believe the GOP-lost-but-conservatism-won meme that's circulating, especially the version that stresses Republicans' apostasy from the '94 gospel of small government and fiscal responsibility, or the one that points to the power of libertarian spoiler candidates. Hell, I hope it sticks if it'll dissuade a few of the remnants from tacking left—or even just make Ryan Sager a few extra bucks in royalties. But I also think primary reason people are trying to push the notion mostly out of a combination of wishful thinking and that same desire to preempt a GOP shift centerwards.
Alas, Ramesh Ponnuru's cover story in the most recent National Review probably hits a good deal closer to the mark. Iraq and corruption are looking like the core factors right now, and I'd wager that for every principled fiscal conservative voter grumbling about the bank-breaking prescription drug benefit, there are two whose beef is that it was too stingy. The complaints about out-of-control spending seem to be a lot more likely to come from ideological pundits—and I think the libertarian-dissent narrative is to some extent appealing to that set because, while neither side's intellectuals are "libertarian" by a stretch, they tend to be more libertarian than the bases they represent (i.e. urban conservative writers are socially liberal as conservatives go; mainstream Dem pundits are constrained by at least a passing acquaintance with economics)
Admittedly, I'm not offering any actual "evidence" or "data" in support of this intuition. As The Wire's Proposition Joe might say, things happen at the polls; proof is hard to come by. But two things kept running through my head as I read the recent Cato study, optimistically pegging libertarians (pretty broadly defined) at 13 percent of the electorate: (1) That's a nice chunk of votes, but still a small enough proportion that you'd gain net votes by appealing to them at the expense of other groups, and (2) People are a lot more prepared to decry "big government" in general than any particular program. The Cato survey question is (necessarily) general, but legislative elections tend to be particular.
A lot of the empirical case for the "covert victory" thesis seems to involve pointing to people like Heath Shuler and Bob Casey (social rather than economic conservatives). But I don't know how far this stretches. We're getting a House swing in the vicinity of 30 seats, which after the Republican Revolution in ’94 is the biggest net shift in a midterm in 20 years. So of course when you finally get flips in districts that have gone Republican for many terms, anyone who’s going to win in these places is going to be pretty conservative. North Carolina is just not going to elect a Barbara Boxer or a Russ Feingold. Other things equal, the median elected official of either party is going to be closer to the center when they’re in the majority than in the minority, when they’re down to their hardcore base. That's just the upshot of the fact that growth happens at the margin.