Originally Posted by Peter Olivola

No, it wouldn't. Between the launch of a small balloon and the launch of the rocket, wind vectors will have changed multiple times. Small Balloon wind determination produces snapshots of a dynamic environment. Spring and fall, especially, experience multiple wind shear phenomenon at quite low altitudes. Those seasons also produce the largest directional and velocity shifts in the shortest periods of time. Any effort to use a balloon to predict winds over a fixed location is an exercise in chaos theory. The change rates for altitude, velocity and direction over a single location will yield little more than a general idea. Fine for hot air ballooning (they are considerably less perturbed.) Not very useful in the context of model rocket flight/descent profile prediction.

IIRC there is someone doing multiple flight profile tracking from a single launch location on a single day using GPS. All the plots I saw showed scattering in excess of 90 degrees and in some cases approaching 180 degrees in both boost/coast and decent.

Just as model rockect engine thrust performance varies widely even within the same batch, so, too, do model rocket flight profiles for reasons having nothing to do with winds aloft (vectored thrust, rod whip, etc.)

About the only useful site prediction tool I've ever seen is a bubble generator for determining when the site is below a thermal. Which introduces another variable that makes using small balloons useful for hot air balloons and useless for rockets. As a thought experiment, use a prediction program to build and test a rocket with absolute minimum stability leaving the launch rod/tower (an Apogee D3 flight comes to mind.) Now lower the velocity at that point in the flight by 10% (the effect of launching into a strong thermal.) Assuming zero wind, how will the rocket react? The answer is, unpredictably (rod whip, vectored thrust, etc., and unknowable.)

Like altitude prediction calculations, there are enough unquantifiable variables in trying to do such predictions that they should be considered advisory at best. The popularity of spot landing competition is a function of its randomness, much like Thanksgiving turkey shoots.