So that's not good because setting a shorter distance on the viewfinder scale would result in even more sky. (If I've got my geometry right.)
If the lens is focused at infinity and the viewfinder is too, their "cones of imaging" are parallel, and the distance between the two is immaterial given the subject is far away.
As the lens is focused closer, the viewfinder must be tilted down by the tilt mechanism so that its "cone of imaging" is now aiming at the subject which has moved closer to the lens. That's the paradigm: as the subject is drawn closer to the lens, the viewfinder tilts further down; shorter distances would result in less sky. That's one reason the cold shoe is often immediately above the lens, so that the viewfinder need only move up and down along the vertical axis to provide proper imaging.
In a similar fashion, when you have moving framelines in the primary viewfinder on the camera, as in a Leica M-series camera, the frames will move down and to the right (i.e., toward the lens) to compensate for the two parallax differences created by the fact that the viewfinder is both above the lens axis and to the left of the lens axis. Some cameras have frames that even get smaller at the same time to compensate for field size.
As has been noted, using an SLR makes all of this parallax issue moot.
