Het NHC gaat uit van een landfall, noemt GFS een uitbijter:

En Florence haalt inmiddels weer 140 km/hr gemiddeld en sterkt weer snel aan.

The fixes from the aircraft indicate that Florence hasn't gained
much latitude yet, and continues moving westward at about 6 kt.
There is no change to the synoptic reason as a very strong ridge is
forecast to build over the northwestern Atlantic during the next few
days.  This pattern should steer Florence west-northwestward at a
much faster forward speed by Tuesday.  On Wednesday, the hurricane
will likely turn northwestward and slow down somewhat due to another
ridge forming over the Ohio Valley.   The most notable change from
the previous advisory is that most of the models are showing a
faster motion in 3 or 4 days, which unfortunately increases the risk
of a destructive hurricane landfall.  The GFS and its ensemble mean
are outliers from the main model envelope, showing the system near
the Outer Banks, while the rest of the guidance is well to the
southwest. While a near-miss isn't impossible given the spread of
the ECMWF Ensemble, there is much more support for a track to the
southwest, especially considering the poleward bias of the GFS thus
far this season. Therefore, the official track forecast is shifted a
little bit to the southwest, and is very close to the
corrected-consensus models.

https://www.nhc.noaa.gov/text/refresh/MIATCDAT1+shtml/092053.shtml