You are definitely right that a system sitting in the Western Carribean Sea in mid October should not sit well with Floridians... The climatologically derived regions for hurricane frequency shows the area, together with the eastern Gulf of Mexico, as well as the area off the SE U.S. Coast have the greatest numbers of occurrences. In addition, most tracks after in situ generation tend to be N, NNE and NEwards. This unfortunately places Florida in a climate favored location to be affected.

The trouble with climate is its an accumulation of past behavior... It doesn't ultimately say anything about what the pattern at hand will command. In this case, however, there are governing parameters in the synoptics that connote a fairly climate-friendly behaved system; which is to say, Floridians have much to be concerned about. All that means is a track close to western Cuba and then NNE toward the West Coat of Florida is a preferred solution when balancing synoptic prediction of major influences to track guidance, at this time. Not absolute - never a certainty, just favored. This will also be suggest below by model depiction.

As far as timing goes... The trick is, does the amplification in the Nation's Heartland penetrate deeply enough in latitude to induce a southerly steering flow? That will be established if said trough incurs upon and then erodes the ridge over the N Gulf. That would tend to initiate a NW motion to Wilma....and then, the subsequent thinking that a NW motion would have to take on N, then NNE becomes entirely intuitive. The devil's in the details as my mentor always used to say.. It is still possible, though seemingly less likely, that the trough to the N never fully amplifies and leave Wilma after only tugging her N - then all bets would be off.

I am a bit uneasy about the global models handling the expected trough in the Heartland, days 5-7. Take the difference between the 06z and the 12z GFS for example: In the former run, a large positive height anomaly evolves near the Davis Straight area (this being relevant because having a positive height departure there tends to signal troughing in the contiguous eastern U.S.), and in concert a powerhouse U/A impulse slices into the northern Ohio Valley and cuts off... This definitely has enough amplitude to grab hold of Wilma a sling her up the East Coast (having crossed Florida), even impacting areas of Eastern New England. Where as, this day's 12z run is much, much less amplified with the westerlies in the Ohio Valley, and this results a much weaker less influential steering field along the East Coast. These types of large scale mass discontinuities across single model runs do not lend much confidence.

The average of the models I have seen,against my personal intuition, suggests "potentially" impacting in western Florida, 6 days out? The deciding factor pulling me this direction would be the interesting agreement in timing (and almost placement) that the GFS has with the GFDL.. Thereafter, it will remain to be seen how Wilma is lifted in latitude, whether N component or more E component; that will come down, again, to how amplified the trough slated to settle into 80-ishW lon really becomes. But...in the end the bottom line is that it is too early to be precise with much credibility. ho-hum, how would a crystal ball help us now?