Polar Analysis from flight logs?

Eric Greenwell wrote:
...
A Google search turned up laser airspeed sensors that, in concept, could
be used to measure L/D directly from the glider. Some of them were good
for the low speeds we need to measure sink rates. So, have one pointing
forward, one pointing down, divide the forward speed by the sink rate,
and ta-da! L/D. It wouldn't matter what the airmass was doing, since the
measurements are relative to the airmass.
...

The real problem then is to determine where should the down pointing
device exactly point. An error just equal to the gliding angle in
the backward direction will give an infinite L/D, and this just
about 1 degree. The direction relatively to the airframe is variable
with speed and is identical with the direction of apparent weight
only during unaccelarated flight, in this case this is also the direction
of the real weight, but we have no mean to detect that the flight has
no acceleration. Although the method is very simple in theory, I doubt
there is some practical realisation which would produce accurate enough
results for being interesting, even with a lot of money.