This is on purpose. This is exactly why we measure both the response time, and take the tracking shot. That way, we cover both the case where your eye stays static, and the case where you follow a moving object.
The response time (what we use to score TVs) is a representation of the blur when your eye stays static. OLEDs excel at this, hence our great motion blur score for them.
We still take the tracking shot to represent the other component of blur: frame persistence (which includes frame rate). This component, while it greatly depends on the framerate of the content, is still affected by the TV. This is useful for games especially, because we often follow moving objects with our eyes. Even for sports it is useful, for example when you follow the ball with your eyes, but the camera stays static (tennis especially). This isn't part of the motion blur score though.