In Nebraska, an east wind is blowing at 10 m/s, and a boy is standing still, flying a kite. In Kansas, it's dead calm, and a boy is riding a bicycle at 10 m/s, eastbound, with an identical kite tied to the seatpost. That kite is likewise flying. Both kites have the same length of string unfurled, and they're flying at the same altitude. At the same instant, both kite strings snap at the point where they are attached to the kites. The wind keeps blowing in Nebraska, and the bicycle keeps rolling in Kansas. There is no thermal activity, and there are no terrain features to deflect the wind in a vertical direction.
1) Which kite hits the ground first?
2) Which kite is the greater distance from its owner when it hits the ground?
In Nebraska:
(A) There is real wind in this case. The kite is tethered and had no motion w.r.t the ground. That means zero KE. When the string is cut, the kite moves immediately away from the kite handler. Its velocity will depend upon circumstance and design of the kite but will be directly proportional to time, and in the simplifying case linearly proportional .
In Kansas:
(B) In Kansas, the kite is in motion w.r.t the ground, because it is being dragged through the air in order to generate the "wind" that keeps it aloft. This means that it has KE, whether you agree or not.
Once the string is cut, the KE will continue to carry the kite in the direction of the cyclist (Who unlike the belt observer, would have the option to stop). After the KE has dissipated against friction of the air, a force retarding motion, the kite will fall to the ground. Its displacement will be some what East (the same direction as the cyclist) of the point where the string is cut, but the distance to the cyclist is indeterminate.
It is approximately, the absolute displacement is the cyclist, less a small amount proportional to the kites KE, and not directly proportional to time, and certainly not linearly so
Notes:
(C) There is the unjustified assumption that the cyclist's velocity will remain the same after the kite is released. This cannot be so for all possible wind equivalents, as they are all to some degree load dependent. This limitation applies to the treadmill belt, and the level of friction to that belt.
(D) The kite or other object, may be modified to make the difference even greater.
(E) This exposes the lack of KE in the treadmill model, and its obvious influence on that result, that is nor present in real wind.
(F) The real-wind observer is stationary and carries no KE (w.r.t the ground if you insist) but not for the real-wind tethered kite.
(G)For the cysclist and treadmill, both kite and observer have unavoidable KE, as there is motion w.r.t the ground.
(H) Motion of the released kite in Kansas, cannot be said to be predictable. It is not acceptable to wish away thermal effects and the essentially random motion of small scale aerodynamic effects. This also applies to the real wind but they will be small compared to the actual and overriding force of the wind itself.
(I) The level of work is also quite different. etc..etc...