the different alkaloids cross from polar to nonpolar at a particular pH, and about 1 pH measure either side of this has to be used to ensure 98% or better exclusion or inclusion of the alkaloid across the extraction. (roughly speaking... it varies a bit but 1 pH point is about how far the curve takes to get to the other side, soluble or insoluble)
ephedrine's pKa is 9.6 so that would be the same as pseudoephedrine
DXM's pKa is 9.2... so that's not quite big enough a difference, although done repeatedly it would have an effect, if 20% pseudo comes across, then do it three times and that's 20% of 20% of 20% coming across, ie 1/125... and left in the base solution will be 20% of the pseudo each time, which you will basify further and extract out... or something. anyway, they are separable by pH but only if you have accurate pH measurement and there would be probably a minimum extra 5-10% loss from having to fiddle with it so much.
if someone ever invents a method of column chromatography that isn't so damn fiddly you'd be able to use that pH and make one compound slide further down than the other and voila...
i might be remembering wrongly, but meth steam distills whereas ephedrine doesn't. does dxm steam distill? (or vice versa - the other way around sounds more hopeful)