Eric,

The vortices at the edges of clouds I am referring to are in the
horizontal, though they typically have a significant vertical
component as well (otherwise, why bother with them). That's why it is
important to view them from directly below. This makes seeing the
horizontal component much easier.

Observed localized rotation at cloudbase (the only place we can see
it) is substantially lower than the rotation rate observed in dust
devils. This is expected. As we go up a dust devil, it expands with
altitude. Conservation of angular momentum alone will account for a
substantial reduction in rate of rotation.

As for the rate of rotation of the entire system, I have never
measured it. All I can say here and now is that it is slow but
observable. Let's, for the sake of argument, say that it moves three
times as fast as the minutehand on a clock, that is 18 degrees per
minute. (Remember, this is at the edges of the thermal. We would
expect increased rotation within a strong core.) If the cloud is 1/4m
in diameter, the speed of rotation is about 2 knots. That's a 4 knot
differential for a left versus right turn, with corresponding turn
radii for a given angle of bank.

Granted, the system is turbulent. And there are additional factors
that might contribute to large scale rotation such as wind shear,
inversion, perhaps even condensation.

For argument's sake, let's say that it does rotate, on both large and
localized scales. What advantage can we take? How can we detect it?
How might we change our approach, entry, and centering techniques to
maximize overall rate of climb? These are the questions worth
pondering.