I don't understand the reference to the paddlewheel rgnmstr. True wind
direction = wind direction (true north) if the boat was anchored at a
point on the water, e.g. not relative to the boats movement or water
current. To achieve this (with the boat in motion), you need the
apparent wind & speed, SOG (from the GPS) and ship heading (true north)
from your flux gate (COG from GPS introduces error) to get a close
approximation, and a little more accuracy can be had if you include
adjustments for angle of heel error, heading error (compensate for mast
& sail induced air flow), a yaw sensor on the mast, etc ... oh yea, and
you need an instrument package that uses this, some like Raymarine wont
use SOG from the GPS.

Interesting paper published that explains true wind, and notes that 80%
of marine vessels calculate it wrong.

http://www.coaps.fsu.edu/woce/truewind/paper/

Better squeeze up for the Brooks & Gatehouse...

I have an open suggestion to Raymarin to incorporate COG as an option,
which would make me happy. The rest of the capability is overkill for
me (adjustments for angle of heel error, heading error (compensate for
mast & sail induced air flow), a yaw sensor on the mast, etc.)