Hi Ian W,
Yes I know all the arguments about being in a windfield so the ground speed doesn't matter - but I had the feeling that it did. The inertia of the aircraft that was close to stationary had to be overcome for the aircraft to depart at (guess) wind of 60 kts at 200ft plus aircraft eventual flying speed of say 60kts meant that the airframe had to accelerate from stationary to 120kts - agreed helped by the airmass it was in. That inertia has to be overcome regardless and that is due to the frame of reference to the Earth/ ground speed of the airframe. So an extreme example but I assure you that the aircraft lost a lot of height in the turn going past at my eye level in the tower. A more gentle turn may have masked the effect but it would still have been there.
Relative to you, the aircraft had to accelerate from 0 to 120 kts (change of 120 kts). Relative to the aircraft it had to accelerate from 60 kts heading North say to 60 kts heading South (i.e. 120 kts change in air speed vector - i.e. the same vector value)
Your estimation of distance is made using the apparent size of a known object (angle subtended at you eye). Your estimation of height is made using the angle between the object and the horizon. As the aircraft appeared to accelerate unusually rapidly down wind, it's distance from you increased unexpectedly rapidly and the angle between the horizon, the aircraft and your eyeball reduced rapidly giving you the illusion of a height loss.
The illusion is similar when observing a "slow" jumbo compared to a faster 737 on the approach when both are flying at say 160 kts. The jumbo appears slower because it appears closer and yet the change the angular position is the same as the "further away" 737.