[Wavytone]

Andrew,

With a scope pointed at the pole (-90 degrees) the misalignment between the RA axis and what you see in the scope is the sum of five mechanical misalignment errors:

a) the perpendicularity (or lack of it) between the RA and dec axis;
b) the perpendicularity (or lack of it) between the dec axis and the and OTA, at the dovetail or OTA,
c) the parallelism (or lack of it) between the OTA and the actual optical axis as seen in the eyepiece, due to the net alignment errors of the optical components. This is simple enough for a refractor, but gets more complex if there are 2 or more mirrors involved (Newtonians, SCT, Maks or cassegrains).
d) the rotational error in the declination circle or encoder (ie does it really read -90 degrees when the OTA is in the same plane as the RA axis ?)
e) the net effects of flexure in the mount and the OTA (things bend and sag, more than you may realise, even in small mounts).

Determining all of these errors is non-trivial, and each should be determined separately. There is an old optical lab method used to check all of these for theodolites, it can be applied for telescopes but it is a bit pointless as most telescope mounts have no provisions to adjust these.

These days instead the correction can be done mathematically, by pointing the telescope at bright stars all over the sky and measuring where the encoders think its pointing vs the actual star positions. After solving a heap of equations using least squares, a computer can figure out the corrections required to compensate for the above effects. Some - not all - of the better amateur scopes have this capability in their software.