Bunch of questions to get to...

1) FSU model data -- it mostly looks like it's a combination of a lot of stuff happening on our computers plus millions from the web hitting it, causing the slowdown. The PSU site isn't as well known and is on a dedicated web machine that is better able to handle the workload. That said, I still highly recommend our site.

2) Beach -- regarding the upper and lower level winds -- no, those upper level features draw their energy from completely different processes, all mostly confined to the upper levels of the atmosphere. Kraig had a good response regarding the daily variation of the surface winds in those areas.

3) Regarding location of the G-IV dropsondes -- I don't have a map available, but they were all in the general vicinity of the storm, particularly just to the north and west. They did not sample the upper low over the Bahamas; it's too far from the storm. Accordingly, most of the G-IV's impacts are found within the first couple of days of the model forecasts.

And a tad bit of analysis...

Latest recon suggests flight level winds are up to about 115kt with surface winds just a tick under 100kt/115mph. They've also reported concentric eyewalls, potentially suggesting that the storm is peaking for the time being. I wouldn't be surprised to see it upgraded to a major hurricane at the next advisory. I can't remember the last time I saw a 130kt projection from the NHC at the 3-4 day time frame; they are usually loathe to go to near cat. 5 intensity that far out. Still, it's a quite reasonable forecast I'm afraid. A potential disaster scenario is being spit out by some of the models, having Dean mostly avoid all of the major landmasses in the Caribbean and sliding through the Yucatan channel on a course for Texas. That's more of the outlier course right now, as a track over the northern Yucatan is more likely, but one we'll closely watch for the next few days.