Delivery back to HK: *not* heading to Hon Chut

As Grant and the Boyz battle the NE monsoon, big waves and currents, they've decided to head ENE as in the first screen shot below from PredictWind (PW).  It's interesting to note the different routes suggested by PredictWind for a "comfort" route back to HK.... (ignore the lines to the East, to the Philippines, etc.  They are earlier routes).

As of 12:01 today Xena is at 11 10N 110 26E, in winds of 20-25knots,
at least according to PW (Grant has reported higher, to 42knots...)

These are the suggested routes by PW, the red dots based on GFS
and the blue ones based on CMC.  That is, they overlay a PredictWind algorithm
on the GFS and CMS data and labelled as PWG and PWC. They are clearly wildly
divergent.  PWG ETA is 13:00 on 7th Jan, PWC ETA is 17:00 on 6th Jan.
(Green pin is where they are now, 12:01pm, 2 Jan)

This screenshot shows the GFS and CMC suggest routes, without the
PredictWind algorithm overlaid.  They leave Hainan to starboard!
ETAs on theses routings are: GFS 7 Jan 14:00 and CMC 6 Jan 23:30

This screenshot shows the currents, so they're in a positive current for now,
with winds on the beam at about 060 TWA.
Note also the strong current down the Vietnam coast.

All these predictions take account of currents as well as winds.  What jumps out is the wild divergence of the suggested routings, from heading due East, to heading back to the Vietnam coast and leaving Hainan to starboard.  If you ask "which is the best", the answer from the guys at PW is "it depends".... If all the routings are similar, then you have confidence.  If they're not, then you pays your money and you takes your chance: ie, own choice.
Note: that Grant has onboard this same PW data, though not the online screenshots like above, as he gets it by email with no attachments and has no internet. From the current track they've clearly decided to take the fair current and head East,  according to the GFS routing, first part