Bookworm has it right.
I could spout on about Inertial Reference Frames, but let me pose a simple test. In a steady wind of 25 knots or so, trim the aircraft (helo or airplane) into a steady state banked turn, constant power, altitude and airspeed. Watch it for a few circles, and try to see any climbs or descents during its circles.
There are none. None.
The downwind turn is right up there with extra gunmen on the grassy knoll, all myth.
When bad pilots see the groundspeed increase, they pull back on the stick as they turn downwind, they then lose speed and stall, and Darwin gets another point.