Wow...check out the eerie resemblence to the '93 storm's track in the 18z GFDL's track forecast for invest 92L (likely the catalyst for the development in the Gulf later in the week) -- pay particular attention the 72-96hr time frame.
HR LAT LON HEADING/SPEED(KT)
0 23.0 75.2 275./11.1
6 22.9 75.9 262./ 6.6
12 23.2 76.6 289./ 7.3
18 23.7 77.4 304./ 8.7
24 24.5 78.1 323./10.6
30 25.1 79.0 301./ 9.9
36 25.9 79.8 314./11.1
42 26.5 80.8 304./10.6
48 27.3 81.6 312./10.6
54 28.0 82.6 306./10.7
60 28.5 83.6 299./10.6
66 28.8 84.1 296./ 5.5
72 29.2 84.3 335./ 4.4
78 29.5 84.8 303./ 4.8
84 29.3 85.2 247./ 3.9
90 29.2 85.1 173./ .8
96 29.6 84.5 63./ 6.1
102 30.1 83.7 57./ 9.1
108 30.4 82.6 75./10.2
114 31.5 80.4 64./21.2
120 33.4 77.8 53./29.2
126 35.7 75.1 51./31.9
The 18z AF-MM5, which, as Scott noted, was the first to hint at future development (a few days ago) keeps this system on the eastern side of the FL peninsula before moving it to the north. The 12z UKMET still shows the cyclonic loop over the Gulf, except that it turns the system back to the NE far sooner, with a landfall N of Tampa along the FL Big Bend. The 12z NOGAPS doesnt show development until it's already exiting NErn FL near JAX.
The bottom line is that, as Jeff and TimV alluded to, this system may not be entirely tropical in nature and looks to take a very similar track to the 1993 "storm of the century." Regardless of whether it turns out to be a tropical, sub-tropical or some other type of hybrid storm, this appears to pose a significant threat to the NE Gulf coast/SE US by the end of the week.