Yeah, it is. And the problem lies in the very mechanisms that you want to be predictable but in fact defy rules of predictability. Even human nature is unpredictable, and we have a pretty good understanding of that. My guess is that we will be able to predict weather patterns before we are able to predict direct evolutionary genotypic change.
Weather is nonlinear, so obeys chaotic dynamics, hence is sensitively dependent on initial conditions. I'd suspect evolutionary processes to be chaotic as well. I'd be surprised if there has not been research along these lines. Anyone ?