Regarding the FSU models...

We're running two versions of the MM5 during the tropical season, one that uses the GFS for initial and boundary conditions and one that uses a combined GFS + GFDL dataset for initial conditions. The latter gives a better initial representation of any features in the domain, like 95L and TD Ten-E in the East Pacific.

The MM5 is a mesoscale model primarily designed for research purposes. The Air Force used to run an MM5 in real-time for the tropics with results that weren't anything to write home about. I would put more stock in the recognized global and mesoscale models such as the GFS, GFDL, and ECMWF than on anything our MM5s spit out.

That said -- this feature off of the Carolinas (95L) is probably our best candidate for our next tropical cyclone in the basin. It'll take a day or so to undergo warm core transition (or tropical transition), but by late tomorrow I expect we'll likely have something of interest. It will likely meander around in the Gulf Stream for the next few days and with weak steering currents, erratic motion can be expected. Interests from the central Florida coast northward need to watch -- but not worry -- about this one over the coming days.