Don't see where you figure they are predicting 95 bu. better than last year. In fact,  I don't see that they are comparing to last year at all. They aren't saying they are predicting a state yield - only in comparison to the long-term averages at those specific sites.

Still, they are giving about a 50% chance that those sites will have yields almost 70 bu. better than average. There is some tremendous corn potential here in much of Illinois. And their model shows 50% chance of beating the average in Iowa by about 13 bu. While you can argue the exact numbers, this study just tries to quantify what we already know. It is going to be a whopper.