If you were counting on Chafee beating Laffey and then Whitehouse for the Republicans to have 50 Senators, then you already thought that the Democrats would win the Senate. Chafee's chances of beating Laffey in the primary have always been below 50% (I think he was a fool not to run as an independent), and Chafee would also be the underdog against Whitehouse in the general (although, admittedly, not as much of an underdog as Laffey would be). And if Chafee did manage to win, he would almost certainly switch to the Democrats if it would give them control of the Senate, so if Chafee wins the GOP needs to have at least 50 other Senators to hold the Senate.
As for Allen, if he is still up by a few points in polls taken while he was being railed in the media for his "macaca" comment, I think he'll win by at least 10%.
And Kean taking the lead isn't the only recent bright spot for the GOP in Senate races; Talent appears to have taken the lead as well, and both Stabenow and Cantwell are looking very weak. And just today a poll came out showing Mfume beating Cardin by 4% in the MD Democrat primary; if Mfume gets the nod, Michael Steele will be a slight favorite in the general election (Steele would be a slight underdog against Cardin).
RI and Virginia in tandem causes the change. I think Virginia is basically a tossup, and that Allen might lose. That pic with Allen with the white supremicists, is going to damage him in the DC burbs area, and elsewhere, I suspect. That combined with the macaca remark, may well cause some to connect the dots as they think appropriate. I don't see Maryland going GOP, unless the Dem candidate performs poorly after the primary in presenting himself. I just don't. So assuming NJ stays Dem, and I think Kean has a real chance, maybe a 35%-40% chance, we have Penn and RI likely lost, and Montana not there yet, but close to it, with Missouri skin tight (I agree that the Dem has flaws), Virginia a real problem, and Tenn somewhat in play - it remains to be seen, I see the Dems as up to about 45% odds in getting five seats. You are right, that getting six seats, that the odds are lower. That might be around 33%. In any event, the over under is around just over 4 seats IMO.
Yikes I forgot Ohio. The over/under is close to 5 seats, GOP loss, maybe 4.75. DeWine might be put as a 40%-45% chance to win.