It must be stressed that there is only a vague indication of what parts of Florida are under high risk with this storm. The latest GFS model guidance (leading U.S. model) shows a landfall in northeast FL around Daytona Beach with a slow northward push over land into Georgia. This would be a high impact scenario for the JAX region.
But the latest European model guidance is considerably further south with landfall around Fort Lauderdale or West Palm Beach and slow northward motion over land through all of central FL. The Canadian model is closer to the Euro than the GFS, landfall approximately Jupiter or Vero Beach area, high impact later around Orlando.
The one general theme overnight is to lessen the chances of a second landfall on the Gulf coast, almost all guidance now shows Dorian inland over the southeast then in some cases near the coast of the Carolinas after several days, probably at tropical storm intensity due to the long stay over land.
As the NHC cone suggests, all of Florida is at some risk from Dorian. As to intensity, with the track being all over warm water from now to Florida and Bahamas not having much if any disruptive influence, cannot rule out any intensity although cat-3 seems most likely from historical analogues.
I would expect a little better consensus on landfall location and timing by about Friday night or Saturday morning. The current spread and uncertainty is caused by a rather weak set of steering currents coming into play by Monday-Tuesday over the southeastern U.S., and in fact some model guidance suggests a very slow-moving storm that might be impacting given locations for 24-36 hours.
Also note that storm surge potential is high with new moon timed for early this weekend.
WHich was the better models in the past? Euro or US?
I am currently betting on the Euro model right now.