Sept. - Oct. can be a very difficult time of the year to forecast recurvature on these storms in the western Caribbean and southern Gulf because of their interaction with strengthening mid-latitude systems. Combine that with the fact that we have very minimal data over water, which causes questions in model initialization ... Ike is a storm to keep a close watch on through the next SEVERAL days. The GFS has been quite reliable this season (as noted by others) and it was already suggesting last Tuesday that Ike would move into eastern Cuba around Sunday.

My question is how much will Ike weaken if it moves across/along Cuba? Intensity and timing are still big question marks. Note that the forward speed of Ike is projected to slow down as it enters the Gulf by Wednesday. IF SO, it will be a prime candidate to make a sharp turn to the NE and the FL Gulf coast by Friday, in response to a deep trough forecast to push into the central US. But that's a long way out yet.

This time of the year (esp. late Sep.-early Oct.), we've seen models try to recurve these cyclones over the long term, only for them to continue moving west, south of any mid-latitude influence. While that's not the likely scenario right now, it bears watching.