Actually, I know what Bryan is saying, and I think people are missing the premise.
Conditions: Calm. Plane facing west, treadmill moving east.
Put a plane on a treadmill, also put a speedometer on a wheel. The only way that the speedometer reads the same speed of the treadmill is when the forward motion of the plane is 0. Any eastward movement by the plane and the treadmill speed will be greater than that of the speedometer reading. Any westward movement of the plane will result in a speedometer reading higher than that of the treadmill. If the treadmill is moving at 70kts, and the speedometer will say 70kts, and the airspeed says 0, the engine is providing just enough thrust to counter any friction loss. However, any extra thrust providing any forward airspeed will result in the speedometer reading more than what the treadmill is moving at - in which case the experiment is now void because the treadmill is NOT matching the speed of the wheels.