Regarding landfalls, FL & GA remain safe. SC is very close to safe but not a done deal up near Myrtle. NC obviously is still at risk, especially the OB. Regardless, the GFS pressures are laughably way too low imo like usual for just off the US east coast just like it was with Jose and also Irma in the FL Straits. Add 20+ mb & that should be closer.
Note that Maria has had pressure only down in the 970s today. Compare that to those ridiculously low pressures the GFS runs of just 3-4 days ago had for the current time (still available at Tropical Tidbits' archived runs):
Run SLP as of 18Z 9/26 (mb)
18Z of 9/22 929
0Z of 9/23 937
6Z of 9/23 935
12Z of 9/23 939
18Z of 9/23 938
Now watch as we move forward and the runs correct for the bias of too low pressure:
0Z of 9/24 945
6Z of 9/24 949
12Z of 9/24 950
18Z of 9/24 962
0Z of 9/25 964
6Z of 9/25 966
12Z of 9/25 968
18Z of 9/25 969
0Z of 9/26 973
In the quoted post, which was made soon after the 18Z 9/22 GFS with its 929 mb for 18Z 9/26 (and 933 mb for the subsequent map of that run), I said, "Add 20+ mb & that should be closer." Whereas that turned out to verify, the best bet would have been to add 40 mb! This is just more evidence that the GFS has a serious low SLP bias for TCs at least near the US SE coast. That needs to be fixed, if possible. Regardless, it is important that GFS users like us learn from verifications like this and keep biases in mind.