One of my goals (and that of many others on the site) is not just to track storms but to teach others about them. Notice that I said 'pose a threat' to Florida, not necessarily hit it, others have passed through the box in the past but have missed the state. You are correct in assuming that normally the threat is greater for south Florida, and you are also correct in that the storm can indeed get caught on the wrong side of the ridge, i.e., its not picked up by the trough but instead gets blocked by the ridge behind it and does not move off to the northeast. Sometimes tropical cyclones can slow down for days before they finally head off to the north northeast on the back side of the western extension of the ridge. When this happens (more often than I care for) the storm can get rather close to Florida - and its still one of a few scenarios that can play out with Earl. Hurricane Bertha in 1996 is a classic example of this. Bertha was expected to turn on every advisory as she approached Florida - but she didn't. Finally the hurricane got too close and so from a safety standpoint east central Florida beachside communities were evacuated - and two and a half hours later she turned!
Here is a link to Bertha - notice how similar the track is to what Earl has been doing.

UNISYS Track of Hurricane Bertha

Cheers,
ED