I've looked over the 10 or so orphaned electrophoretic variant genes in the Dmel genome and only 3 were enzymes and were sufficiently localized to attempt to correlate them with genes with relevant GO terms.  In two of the three cases, there are two adjacent genes that could correspond to the electrophoretic variant.  In the third case, there appears to be only one candidate.
These are:
Gene    Location     Candidate Gene(s)    Location(s)
Agi     33A1-33B2    CG14934, CG14935     33A3-33A4; 33A4-33A4
Fum     5C1-6C11     l(1)G0255, CG4095    6C5-6C5; 6C5-6C5
Mdh2    90E2-91A2    CG7998               90F8-90F9
Bill Gelbart
FlyBase
Harvard University