Air movement is a rounding error in this test.

Suppose the curvature of the wing's airflow could in fact push the drone 15cm up or down in the time it takes the wing to advance 30cm forward (a tiny fraction of a second). That would require such a high G load exerted on the drone, it would disintegrate even if it didn't hit the airplane.

But of course, the air density and curvature is not nearly enough to affect a drone so dramatically. Mooney M20's wing generates only about 15 pounds of lift (down force) per square foot of the wing at cruise speed.

If instead of the drone we had a feather (small low density item) and the speed was 5 knots then yes, that might have worked.