One thing I've wondered about in this direction, is something similar. But with the use of pthalimido protecting groups, whilst not going into details the synthesis of pthalimidopropiophenones is not difficult. Its actually an intermediate step in the Gabriel synthesis, after which, whatever target compound whoever is doing whatever, that pthalimide is cleaved off with hydrazine (fairly nasty stuff) The synthetic details are relevant here, not in detail, but the fact that these pthalimidopropiophenones are intermediates of that nature, well known, well characterised etc. as a class (I don't mean specifically of the 2,5-OCH3-beta-C=O phenethylamines, but phthalimidopropiophenones or other such well known substitution patterns on a BK-amphetamine)
The primary keto-amines undergo a dimerization and cyclization giving diphenylpyrazine products (how delightful)
To dimerise successfully to a significant degree they are going to have to be present in a concentrated enough solution, thats going to be a bugger if one was trying to work up a synthetic reaction and isolate the end product substituted cathinone (or cathinone ltself) without ending up instead, with a mess of the corresponding diphenylpyrazine, some cathinone and rxn byproduct shite, but if the concentration is dilute enough, there are not going to be enough molecules meeting to do so. The phthalimide protecting group gets cleaved off in-vivo, giving the substituted cathinone, to be diluted in stomach acid and absorbed as it does so.
Pthalimidopropiophenone itself turned up in some variety (or varieties, I don't know) of party pill, as a prodrug for cathinone. I see no reason why the principle couldn't be extended to pthalimidopropiophenones bearing the (for instance) 2,5-dimethoxy-4-R-phenyl-beta-carbonylpropan-2-amines, (R=halide, alkyl, alkoxy, thioalkyl, etc. or for ex. the 3,4,5-same.
Thing is...the body will do the same thing