Cindy moving northeast should help to erode the western extent of the ridge, not really to build it up. Were it moving due north, or accelerating off towards the east, that'd probably be the case. The trough trailing behind the storm coupled with another shortwave trough upstream of the system are & will more than likely help to erode the western extent of the ridge.

The models have really been going back and forth as it is a question of small-scale features and their timing that will influence the future track of the storm. Any small change in one or the other could result in a large forecast track (and to a lesser extent, intensity) change. I wouldn't pay too much attention to the models out to 5 days now other than to say that everyone east of Texas needs to be watching this one very closely.

Question about the GFDL/GFS: I do believe the GFDL gets the initial conditions from the GFS, but the forecast itself comes from its own model. The storm itself is initialized separately using a bogussing technique. Most US mesoscale/fine-scale models are initialized using the GFS as a background, for what it's worth.

To better answer the convection/structure question: my gut feeling is that the storm has ingested a bit of dry air coming off of the slopes of Hispaniola, affecting convection on the northwest side of the storm, but this is only a temporary impact. As it begins to move a bit further to the west, as we are seeing now, the convection should wrap entirely around the center and allow the storm to resume strengthening. Truth be told, any hold up in intensification is a good thing.