While I have no experience building track, I do have some engineering experience (NOT railroad engineering...mechanical & materials engineering). I don't see any connection between issues of expansion and contraction and the issue of soldering joints. I may not have thought of something, but to my mind, whether a section of track is soldered or not has nothing whatsoever to do with the track's behavior through temp changes. Expansion in either case does not happen at the joints--it happens along the entire length of each piece of track.
The only difference in the two situations that I can imagine is this: if the joints aren't soldered, then the individual sections of track can expand toward one another until a) they physically butt up against one another, or b) the screws or clips on the rails resist any further relative motion of the rail. Case b) is equivalent, at that point, to case a).
When cases a) or b) have been reached, the situation is identical to rails that are soldered together. At that point, the entire track shape itself expands (or contracts).
This all from a guy who has never built yet a single track. Tell me where I'm wrong.