?? well, one source of confusion about possibility of fitting gl(2,f_3) into mckay correspondence is ... ?? "binary tetrahedral group" (=?= "e6" ???) only covers _rotations_ of tetrahedron .... ??? .....

?? on the other hand, what about e7 ?? ... ??? still doesn't seem to have right number of irreps .... ?? what about number of even irreps ??? .... ?? still seems wrong ?? .... ?? ....

?? hmm, maybe extra dot is key here ??? ... ?? neighbors of extra dot as "spinor" rep ??? ....

?? e7 with extra dot .... ??? ....

?? ok, this seems like it might make sense .... ?? also mesh with child's drawing stuff ??? ..... ???? .....