I don't have a lot to add to what you already know about the results of last night's elections. I basically agree with Jonathan Chait that it was a good night for the Democrats. Of the four possible matchups in Kentucky, I think this is the one that gives the Dems the best chance to pick up the seat; I'm not certain that Joe Sestak is a better general election candidate than Arlen Specter, but I agree that the odds are that he is. Winning the special in PA-12 is obviously nice for the Democrats. The one sour note for the Democrats -- the need for a runoff in the Arkansas Senate primary -- would be, in my view, a negative, except that I don't think anyone expects the Democrats to keep that seat anyway.
Which points to the sort of double context that these things need to be placed in. On the one hand "good" news for Democrats would be indications that they'll only lose historically normal numbers of seats this year: 20-30 in the House, a handful in the Senate. So that's not really good, is it? But on the other hand, Republicans can make that sort of gains and still fall well short of taking control of either House of Congress. If the GOP picks up 25 House seats and 5 Senate seats, and Democrats control both bodies, then declaring a winner on election night will be about spin and semantics.
Beyond those issues, I probably agree with what I believe is the conventional wisdom. Yes, it's a bad environment for incumbents, as the New York Times headline this morning says., although of course incumbents will win the vast majority of the races they contest, as usual. Still, two Senators defeated for renomination (so far, with Lincoln in Arkansas hanging by a thread, and McCain still in trouble in Arizona) is unusual.
There's also the general sense that Republicans may eventually be harmed by their inability to nominate electable candidates. To me, that's the story of Rand Paul. Of course, Paul may well win the general election, but I continue to think there's a good chance that Republican gains this year will be harmed overall by the nomination of ideologically extreme nominees, and in some cases less capable candidates, and by the pressure in other districts for mainstream conservatives to act as if they were ideologically extreme. It will be interesting to see how Paul in particular fares in a general election context; Kentucky is a good state for Republicans, and with a mainstream conservative candidate I don't think it would have been a contest, but now I'd expect a fair amount of uncertainty. The question is how many districts around the nation are having similar results. Hey, reporters! More about 2010 House nominees, please! (And by the way, reporters -- I'm guessing that at the lower levels, state legislative races for example, you'll find some great stories of very fringe candidates getting nominated).
So, while I wouldn't make too much of any one set of primary results: good day for the Democrats.