The GOP defeated 34 DEM incumbents. Of these:
5 were in 60%+ GOP districts
8 were in 55-60% GOP districts
4 were in 50-55% GOP districts
8 were in 45-50% GOP districts
7 were in 40-45% GOP districts
2 were in 39%- GOP districts
In fact, IL-05 was about a 33% GOP district and NJ-08 was about a 36% GOP district. More to the point, of the 34 Dem incumbents that were defeated, exactly half were in Dem leaning districts. 9 were in districts that were less GOP than the current CA-50 district is Dem.
Of the 22 open DEM seats picked up by the GOP, here was the partisan-lean breakdown.
6 were in 60%+ GOP districts
3 were in 55-60% GOP districts
7 were in 50-55% GOP districts
3 were in 45-50% GOP districts
3 were in 40-45% GOP districts
So, although the open seat ratio was more inherently GOP leaning, there were still 6 Dem leaning districts picked up by the GOP. The 4 open seats that were picked up by the Dems all had a strong Dem lean except for MN-06, which was slightly Dem.
And many of these seats were picked up by second or third tier challengers. Of course many of the Dem-leaning seats reverted to the Dems in subsequent elections (14 of the 23, to be exact), but for at least 1994 they were part of the GOP wave.
The reason that I was curious about the 1994 breakdown was because a great number of the competitive GOP seats this year have GOP leans in the low 50s. So, I was wondering to myself if these type of seats (in reverse) did indeed fall to the GOP in the 1994 wave, and sure enough they comprised nearly half of the GOP pick ups of that year. In fact, a good number of Dem-held seats fell to the GOP in 1994 that were more Dem-leaning than just about any competitive GOP-held seat this year is GOP leaning.
It's also worth noting that several of the huge 1994 upsets weren't even on the radar screen in May of that year.
If the 50th turns out to be about an even match, either way, that would suggest an erosion of 5% to 7% in the GOP vote versus what Bush 2004 got. (Maybe the South will remain a different playing field, somewhat, but that is by no means clear, and in any event, most of the action is not in the South.) That suggests that in all currently GOP held seats that where there is a creditable Dem challenger, and/or where the Pubbie is damaged goods, that went for Bush 2004 by say 55% to 57% or less, that the seat is in serious play. (Sure there are some oddities like Leach of Iowa.) That I suspect is about 20-25 GOP seats. Except for the Ney seat, that probably includes all your GOP seats that are in the lean category or less, plus a few of the likely GOP seats.
What is your partisan lean baseline by the way? That is a rather critical question come to think of it. Bush was such a good partisan baseline. It was much messier in 1994. I say that because NJ-8 was not nearly as Dem as you suggest, generically. IL-5 was Rosty, no?
Thinking of 94. Let's take the converse. How many races that were definitely on the radar screen as being Dem vulnerable did the Dems end up winning. You guys think about it. In a sec I'll post my recollection.
The IL-05 was carried by George H.W. Bush by 52%-48% in 1988, and the NJ-08 was comfortably carried by Bush in 1988 and would have been fairly evenly divided in 1992 had Perot not been on the ballot. The GOP did not pick up any House seats that were as heavily Democrat in presidential elections as the CA-50 is Republican in presidential elections.
The RATs can't pull off a 1994 this year because there just aren't enough RAT-leaning seats held by Republicans. Even if the RATs win the CA-50 in a low-turnout special election caused by the GOP incumbent's conviction for bribery, that won't be a harbinger of anything.