How about this "bungee effect" scenario:

You accelerate the glider smartly and stretch the rope. The glider lifts
off normally but encounters a strong headwind layer at 100 feet AGL. The
pilot frantically signals for less speed and the winch driver promptly
reduces power, but the "bungee effect" just keeps pulling the glider faster
until the stored energy is depleted. Then the glider airspeed sags as the
reduced winch engine power is felt. The pilot signals for more speed and
the winch engine throttle is again advanced, but the additional power is
absorbed in stretching the rope before the glider sees the additional speed.
Then repeat cycle.

I once tried auto tow with 3000 feet of 5mm nylon and this is what I
encountered. We never did tame the launch so we went back to steel wire.
_______

This is a very bad way to control a launch unless the winch power is set
very low so the glider pilot controls airspeed with pitch attitude. I don't
like the idea of winch rope that stretches any more than needed to damp
vibrations. I like the idea of instantaneous airspeed response to throttle
changes. Spectra stretches less than 1% at the failure point. Spectra is
pretty much a "drop in" replacement for steel wire except that it is 90%
lighter.

It's much better to have reserve power from a large, high torque engine and
a winch line with little stretch. My dream winch is a powerful one that has
a telemetry link to the glider. The winch pulls the glider to the desired
airspeed at the maximum safe acceleration and holds it precisely there while
the pilot pulls up to tension the rope to 80 - 90% of the weak-link breaking
point. That is the definition of an optimized launch.

Bill Daniels