This is widely claimed, but it doesn't seem to be true. I
found that hot and room-temperature naphtha both pulled gramine in close proportion to 5-MeO-DMT. Gramine is depleted somewhat vs. DMT, but probably by a factor of less than two with two pulls, less with more pulls. Perhaps the result would be different at higher concentration, but I ran my experiment around the same concentration that I've found convenient for larger-scale extractions.
If you can get the alkaloids to crystallize, then the major constituent (weighted by solubility) will usually be purified. So if the major constituent is a tryptamine then you'll get that, but if it's gramine then you'll get purified gramine. There is no way to judge without TLC or other analysis. The color does not provide an indication, since gramine is colorless in solution; the color is other impurities. I confirmed rejection of gramine and DMT when I crystallized 5-MeO-DMT oxalate. I couldn't get my extract to crystallize as freebase but others report success, I guess just with different genetics.
All that said, gramine doesn't seem to be particularly harmful. I strongly suspect that most adverse experiences with
Phalaris are a 5-MeO-DMT ovedose (which I again note has probably killed at least one person, so take care; the risk of oral use with harmalas is particularly high).