While the GFS is the preferred model of the NHC, it's also demonstrably less accurate than the European...certainly, I want to be wrong here, but I'm going with the sources (people and models) I have trusted over time. They expect it to hit land somewhere near Jupiter. It actually would be a bit worse overall should that landfall be delayed a bit (say, Melbourne or Cocoa) as the winds would stay stronger over a wider area.
Meanwhile, even the NHC now indicates a landfall at 8am Friday at Titusville/Cape Canaveral.
it's also demonstrably less accurate than the European..Yes, the European models are usually more accurate.
Seen the latest? Matt loops around and crosses S. Fl and the Keys. Crazy.