Below is the NWS Ocean Prediction Center's 24 hour forecast. You can see the high pressure area to the north acting as a blocking mechanism moving eastward with a high pressure area moving slowly westward in the GOM as well. With this progressive pattern, a trough pushes eastward from the northern plains, degrading the ridge to the north, eventually allowing 91L to escape in that direction. This is what looks to create our classical loop setup.
The 12z ECMWF run from yesterday was very slow with moving this trough eastward, which allowed 91L to remain in a weak steering current longer and created a more direct impact on the U.S. as well (which was an outlier compared to the other models and even had little agreement from the ECMWF ensembles). It appears to me the 00z ECMWF run has sped up the approaching trough a bit more, pushing that high pressure area to the east quicker, which in turn allows 91L to move towards the north quicker (with no direct U.S. impact). The timing of the trough and how quickly it breaks down the ridge to the north will determine when 91L will be able to move north and then northeast around the periphery of that high pressure area. Nonetheless, it's a very complicated setup for models to forecast and results in weak steering.
Also, the ECMWF has a high pressure area moving eastward behind the trough that is expected to pickup 91L and move it out to sea. Generally, when we have this setup (and Alex from 2004 is a good example) we have a high pressure area to the east and to the west of the system, and we see a large pressure gradient that allows for a system to strengthen (especially if this system were to stay over the Gulf Stream, in a low shear environment). At the time, the strengthening of Alex was unexpected and not forecast by the models (if memory serves correct).
