Again, the polls seem to have some strange results most notably being Hawaii and Wyoming. While they are both safe seats, I doubt that the margin has ballooned itself to 54 and 58 points. The YouGov polls seem to have the biggest problem with minorities and the ability to gauge the exact size of large leads. (the latter of which is a very common problem)
Notice that all of the races where the trailing party has a >20% chance of winning are seats in which the GOP current leads. Meaning that unless things change in NC, MN, NH, or MI the GOP is currently at it's ceiling. Since I assign Orman to the majority party, a turn of events for Roberts in Kansas would still leave the Senate at 53-47. This is why the average seats in my simulations is still under 51. The mean is under 51 while the mode is 52. The GOP chances climbed another 2 points today.