OK,
Scenario B2 (really fast running person inside airliner. He runs at 501 knots and can turn around on a dime.)
final groundspeed minus initial groundspeed
(airspeed + tailwind) - (airspeed - headwind)
or, if you like:
(walk + tailwind) - (walk - headwind)
(501+500)-(501-500)
1001-(1)
1000
Scenario B3 (really really fast running person inside airliner. He runs at 1000 knots and can turn around on a dime.)
final groundspeed minus initial groundspeed
(airspeed + tailwind) - (airspeed - headwind)
or, if you like:
(walk + tailwind) - (walk - headwind)
(1000+500)-(1000-500)
1500-(500)
1000
Holy schnike, it turns out that if you substitute any number for the airspeed, the groundspeed difference is 1000, which is twice the wind speed when going from upwind to downwind.
Unless that number is less than the wind speed, then it becomes a special case for some reason that wraps people around the axle over scalars and quantum chromodynamics and whatever else.