There is no such problem because you are shooting at objects very, very far away.
If you take a picture of a mountain range, you can move a few steps sideways and it won't change anything to the picture if the camera has the same direction.
More scientifically, the pixels in the picture represent the amount of light with a given incidence angle on the lens. The tracker compensates the earth rotation so the lens axis keeps the same absolute direction in space. Stars are so far away that as long at the lens axis is the same, the change in angle of incidence due to a position shift of the lens in space (arc sine of the lens shift over the distance of the star) is very, very small.
In fact the biggest shift doesn't come from the rotation of the camera around the mount axis, but from the movement of the earth on its orbit (30km/s). You would start noticing a difference for close stars only if you took pictures several days apart (because the camera would have moved enough, this is how the distance to close stars is measured by astronomers)(*).
(*) This is where the "parsec" astronomical unit comes from.The closest stars are more than one parsec away.