Pretty sure the answer to that is L-(+)-tartaric acid which is also used a lot in brewing / winemaking. It is a substance used for separating isomers of various compounds, also like amphetamine for example.
As long as there is a significant difference in how (energy efficiently) this tartaric acid isomer forms salts with the isomers of a drug, you should see that as a difference in tendency to crystallize... allowing you to crystallize one isomer of the drug first and the other one after that, after as in under changed conditions like cooled or evaporated or further addition of a second solvent.
Depending on how big that difference is though, and how well you are able to use the correct solvent systems etc... achieving that separate crystallization may range from doable to quite difficult.
What I don't really know is which MXE isomer would crystallize with L-(+)-tartaric acid, I suspect that it's the levo with the levo for point-symmetry.
You don't use LC or GC-MS for chiral separation I don't think, as the physical properties involved with that are not different for the isomers? A polarimeter though can tell you the optical activity.
EDIT: oh the dextro was used... well I guess same story applies, only the levo should be more available?
I wonder, was there a marked gap between when one isomer crystallized and stopped crystallizing, and then the other under changed conditions? Or was it more continuous? If continuous, I guess it would be very hard to really be useful.
Anyway.... with most drugs who would be bothered to isolate the solely active isomer?... but with something like amphetamine or certain dissociatives where the other isomer also has activity but different... then you more or less just have two different drugs that may be worth experiencing by themselves! Something possibly overlooked with analogues of these substances, to be fair we probably often just don't know about the different activity with novel analogues.