We're stretching the outer limits of the models by looking at things 144 and 168 hrs out. Any model that far out is really not very practical for storm forecasting anyway. Looking that far out, one has to assume climatology and the general setup at the end of the model period. All of that suggests that potential Wilma would be heading NE or NNE by that time. So take your favorite model, and extrapolate out the general area from the end of the model following the general path given from the preceding 48 hours.

Basically, if this thing develops, one would expect it to head perhaps a bit west, then N then NE. That puts the eastern side of the Gulf Of Mexico most in play, with Florida being the highest odds. That said, the odds aren't all that great of a Florida hit at this point, the odds are just better for Florida than anywhere else right now.