The FSU MM5 is chugging through it's 0Z run right now (it's up to about 9hrs out last time I refreshed it), and it looks so far to be edging it a bit southward of where it originally had it in the 12Z run before; ie, for it's position at 77W-78W, it looks a whisper-hair farther south in the new model run as the last.

The only downside is this model takes some time to do it's run (ie, it's only just now output the 12hr data as I've typed this.)

I do note one thing: it really predicts a hammering by Cuba on the storm's power: by about 9hrs out it models pressure to rise up to around 987 and wind speeds to drop to around 62kts. That's a heck of a walloping, and I'm doubtful it's going to take that much abuse.