All of this discussion has related to side focus scopes, some of the discussion has questioned the influence of the scope's adjustment mechanism. It's important to note that many front focus scopes also suffer from variable ranging in different temperatures. I had a front focus Leupold which moved a huge amount.
One of the most stable scopes of all, the Burris, is front focus.
The fact that both front and side focus scopes can suffer from shift (and not suffer) might mean that the adjustment mechanism is not solely responsible for shift.
It would be interesting to mount a Schmidt on a one-piece steel mount with the rings as far apart as physically possible and see if the restriction of the scope tube reduces ranging variation. Might be distortion problems though, but I guess that's a separate problem.
My Deben ranges short in cold weather, close to freezing a 55 yard target will resister as about 51-52. Past about 8 or 9 degrees if it varies at all it is by an amount so small that I cannot detect it. For what it's worth, though I'm getting on, my vision is perfect.
My suspicion is that the level and nature of coupling of the lenses to the scope body could be important. In my Deben the front lens carrier is entirely separate from the objective bell and the carrier may not even be alloy.
If in general the lenses which move to adjust focus are mounted on common carriers of say brass then they might not fully experience the expansion and contraction suffered by the scope's alloy body.
I'd say all in all it's largely related to expansion and contraction of various bits of the scope with temperature. The root cause of the problem could well be different in different scopes, and the ones that range more or less the same might just have behaviors that cancel each other out in rangefinding.
I totally accept Andras' test results on the Schmidts, but it really takes some working out how some can jump and others move gradationally. I wonder what they changed in the construction of the scope?
If we knew what Schmidt changed back in 2010 we might at least know what part of the scope is moving during temperature.