That is certainly one of the most epic first posts I have ever seen. I'm guessing he used the same method Chriskang used eons ago (i.e. dwelving into the depths of the game's code, getting the algorithm, calculating and then testing the results). If this is right, the algorithm has been changed a little, because these results aren't the same Chriskang got.
If I were to say how many repicks towers get, I' say it's still or around 10, as my calculations placed 0.01% as the chance of winning a tower from a half-tower deck, which are close to those of FGs (some more than others).
Oh, I went ahead and calculated the mean chanced of winning a card, yielding an impressive 41.25% (the last time we heard about this from Zanz, he talked about a mean chance around 33%).