Lake Toho pointed out the the NOGAPS has been consistent with a "south of Tampa" hit for the last 3 days. If the GFS and GFDL are shifting further north...and yes, Stuart would be further north) then I think we need to pay attention. I believe that those are the 3 most reliable models. I think I remember a storm last year (or it could have been this year) where the outlier was actually the CORRECT model; so it's happened before.
Look at it this way: consistency is the key. We have one (which has been called the outlier) model that has been consistent for 3 days, maybe 4 now. The other two were in line until a fluke yesterday, but the runs are now coming into line with a further north movement...not sure if they have been doing that for just one run or two.
We have seen too many times where the track has shifted a little left, then a little right, then back again to trust any forecasted path right now. That's why we're in a "cone".
If I remember correctly, Ivan was supposed to hit Jamaica, but missed it; hit Cuba but went below it, we all know what happened. And, as alluded to by a prior poster, look at what happened with Rita...it went from hitting Mexico to making landfall in Port Arthur, Texas.
To rule out any portion of the state is ridiculous. Right *now*, this very minute, some areas are in more danger than others. But we are 3 days out and to say that any other areas in FL can "breath a sigh of relief" is not only silly, it's downright irresponsible.
IMHO, I think that it will be a very close call for Cancun and Cozumel, but I don't think it's going to plow into the Yucatan Peninsula. I think it will graze it and then move north, then NE. I don't see a scenario which would have it making a very sharp right hand turn (ENE). I'm not sure that is possible with a storm as large as Wilma.
Please keep in mind, that the above statement is only my opinion, not a forecast.