Hi Thomas,
I've never tested a Richard Lange or certainly a Richard Lange PLM for accuracy "in the lab" or "on the wrist", but I imagine that you'd find more sample variation in rate consistency across positions and states of wind with either model than you would consistently better isochronism from one to the other. What I mean is, any potential performance gains would not be addressing the largest sources of variability from the perspective of the end user and so would likely not be noticed. This is just conjecture on my part, but we're talking about watches that are capable of accuracy on the order of a few seconds a day anyway so...
_john