I'm not sure I agree with this. I know what you're saying, and with truly random data (flipping a coin or rolling dice) you would be absolutely correct. Getting 99 heads in a row has absolutely no bearing on the probability of getting it again on the 100th try. I would question if this data is really random though, as streaks (especially over several games) for many hitters seem to be largely dependent on their confidence and timing at the plate. In some ways, I think you could consider this "rolling with slightly weighted dice". Would make for an interesting study.
I'm not saying that the decision to pitch to Bautista in this case wasn't correct, as I think it was. The increased chance of the Jays winning from an EE walk probably outweighs whatever advantage Jose may have had (if it does in fact exist). I'm also going to throw out the "3 guys in a row" walk portion of the original statement for obvious reasons, although in some ways it is the pitcher equivalent.