Quote: It looks like the models have basically split with the official forecast down the middle.
When the models have a wide divergence, the NHC generally has little choice but to either pick one of the tracks, or split the difference (go with the consensus). That's why you have to look at the cone, rather than the actual forecast lines.