I completelly agree with your general sentiment. That said, there are a not insignificant members well to the east of of main cluster. With that, I'd tend to view Tip's "correction vector" eastward.
note: not panicing--just observing.
Hopefully, the late-to-the-party shift of the GEFS is indicative of a final solution that's a little less tucked than the EC. Many more runs to figure it out.