Have a friend doing the SHTP to Hanalei, and am routing him for practice, using the polars from my boat (Bermuda 40) with a scale up to account for his waterline (Garcia Passoa 47) 

Today, updating with the weather from both NOAA and Predict wind, (GFS & PWC) found this divergence in the Isochronal vs Grid models: http://imgur.com/z5wMr2a

Grid model is basically running the Great Circle Rhumb line, and Isochronal advocates hunting to the North. 

A few days ago, both had showed an initial best course to the south (which was validated by those boats that didn't get becalmed at the Farallons) followed by an arc to the north for perhaps the last 800 miles. 

1. Is this due to my settings? or is it an artifact of the isochronal being "fooled" by a marginally (11 hrs in 10 days) faster course that might disappear? 

2. Tried  ensemble routing after downloading all the Predict Wind choices, and only had PWGEM and GFS calculate. 

3. Noting some proxyserver issues through corporate firewall, (not an issue, just information)