Normally NHC will use a minimum of 6 hours for track forecast guidance (often more) and they do not like to make too significant a track change from one forecast to the next. Since persistence is the most reliable forecast tool in the tropics, this approach can usually work to your advantage - but not always. Case in point was Hurricane Charley heading for Tampa when it was rather obvious that a remarkable late season summer front/trof was going to turn Charley to the northeast.

The CLIPER has actually shown the best model performance on this storm - unusual, but it has. Except for poor intensity initialization, the GFS has also done a good job with the track forecast. The 18Z GFS run actually picked up on the current motion. The 00Z run is now coming in - - it places the storm near 25N 61.5W in 48 hours, then drifts it north for the next four days after that - with the system still below 30N six days from now. Bertha will probably still be around at this time next Wednesday.

Best guesstimate on position at 04Z was 23.0N 56.4W. Thats a motion due west at 12 knots over the past 4 hours and suggests that the ridge to the north has been reinforced.
Cheers,
ED