I wouldn’t rely so much on the 06Z/18Z guidance, which does ingest the full range of data that are incorporated into the 00Z/12Z cycles. If anything, if one compares the latest 06Z guidance vs.
yesterday’s 18Z output, the consensus has actually trended upward; most of the members at 18Z showed a borderline TS/hurricane on 25 August, while the 06Z output has trended toward a solid hurricane or even low-end Category 2 on that date, shortly prior to possible interaction with Hispaniola, which is far from certain and probably less likely to occur, given the faster-than-forecast rate of short-term organisation currently ongoing. Oh, and for once I wouldn’t give either the operational GFS or the entire EPS suite much weight in terms of intensity, for reasons others have already enumerated.