Well, if you're looking at muscle soreness being generally associated with muscle injury (there is some debate as to how closely they are related), then the evidence is fairly clear that, in vivo, physiological anaerobiosis, or physiological acidosis (in other words anaerobic metabolism and/or acidosis in situations one would experience during exercise) do not cause muscle injury, and hence muscle soreness to any great extent. The critical element to most, if not all, muscle injury models is eccentric/lengthening muscle actions (E/L), which, again, in and of themselves don't cause much injury. The initial minor injury elicited by E/L is minor, but sufficient to initiate a cascade of events leading to a secondary, and more substantial injury. The secondary injury is primarily a result of the inflammatory response instigated by the initial lesser mechanical injury. The adaptation the OP cites
"For example, I do a 2 hour run and my quads are sore. I then recover and the next time I do the 2 hour run my quads are less sore or not at all."
is well acknowledged in the scientific world (if you believe that sciency stuff) and attributed to some adaptive response that appears to attenuate the inflammatory response, hence reducing the secondary injury. It certainly wouldn't have anything to do with anaerobic metabolism since, for a 2 hr run, the contribution from anaerobiosis is essentially nill. Anyway, the nature of the adaptive response is still unclear, but happens within a couple weeks after the initial novel exercise, and lasts for quite a while. It doesn't depend on aerobic adaptations (e.g. mitochondrial/capillary density) whatsoever.
Props to you for stating that you know enough to know you don't know much. Most of us are in that boat, but far to few are willing to admit it.
Steve
http://www.PeaksCoachingGroup.com