0Z Euro. trend very very slightly west
Late tonight, over or very near the Turks and Caicos
Landfall, Maimi proper Sunday Morning. Cat 4/5, exits coast near Fort Pierce late Sunday afternoon
Clips Cape Canaveral Sunday Night Cat 3
Another landfall in GA Midway Between Brunswick and Fernandina Beach, FL cat 3 late Monday morning. Stays Hurricane into Georgia until just south of Athens (Milledgeville, GA) Early Tuesday morning.

Note: Euro Run ends with Jose as a major In the SW Bahamas (over andros) approaching Florida, after doing a Jeanne style loop Sep 17th.

HMON mimics the Euro run, Landfall near Mimi Sunday morning, Cat 5, exit near Ft. Pierce hugs coast, Landfall Brunswick, GA Cat 4
HWRF closer to GFS, slight shift east, although not as dramatic as the 0z GFS Cat 3 landfall just north of Charleston, SC.

6Z GFS shifts west 30 miles or so closer to Florida this run, still with the bad pressure init, It goes over Turks and Caicos and crooked Island, and the south tip of Andros Island early saturday morning also. landfall near Charleston, SC late Monday afternoon as a Cat 5.

The 5am National hurricane center discussion says it a bit better:
"The UKMET, UK Ensemble mean, and the NAVGEM are the models showing the latest turn, and they forecast Irma to move into
the southeastern Gulf of Mexico and near the west coast of the Florida Peninsula.

The ECMWF and ECMWF Ensemble mean are in the middle of the model pack and show Irma moving over the southeastern portion of the Florida Peninsula.

The GFS, Canadian, and GFS Ensemble mean show the earliest turn and show Irma moving east of the coast of Florida toward the southeastern United States. The new forecast track will best follow the ECMWF, as well as the Florida State Superensemble and the HFIP Corrected Consensus, and it calls for the center to move over portions of the southeastern Florida Peninsula between 72-96 h and then across the Atlantic into southern South Carolina by 120 h.

Users are again reminded not to focus on the exact track since the average NHC track errors at days 3, 4, and 5 are about 120, 175, and 225 miles, respectively."

That last line is very important.