Okay, well its a revolutionary concept, the content however is imperfect.
For one, I highly doubt Schiff will win the CT GOP primary. If he does, Dodd will survive.
There is no way Burr or Blunt is losing in a landslide of this proportions. Whatever there issues are, they would be carried to victory on the slant of the year alone.
New York's class I seat is not going to be lost to anyone except maybe either Rudy or Pataki and definately not to some mayor from some small town. If the Mayor of Yonkers couldn't get 35% against Hillary Clinton, a smaller town mayor isn't getting over that either.
Next, there has been no primary polling in Nevada, that I have seen so far and my guess would be that Sue Lowden would have a slight advantage over Tarkanian but since that is in the air I won't criticize your call on that.
Lastly the polling we do have in Kansas says that Jerry Moran is leading Tiarht narrowly with most of NE Kansas undecided and they will decide the winner. Moran leads the Western part of the state and Tiarht in the SE portion. I think Moran would have the edge in NE Kansas but in such an instance I think for credibility purposes alone, you should go with the current polling data.
Finally some of your margins are bit out of left field.
Other then those issues, its great and again a revolutionary concept.