as far as characterizing the results of a bio synth experiment on vanillin, i would think the usual polar-non-polar workup follwed by vacuum distillation would work for seperating unreacted vanillin from 3OH-4MeO-phenyl-a-hydroxy acetone, if their boiling points are somewhat seperated.
you could then use some sort of paper chromotagraphy or other similiarly lowtech procedure to check that the fraction that is supposed to be 3OH-4MeO-phenyl-a-hydroxy acetone is indeed just one substance.
if their boiling points are too close, then to get a quantitative test, youd proobly have to vacuum distill vanillin/3OH-4MeO-phenyl-a-hydroxy acetone fraction, then you could perhaps find conditions that would selectivly oxidize the vanillin to vannillic acid, which could then presumably be seprated from the 3OH-4MeO-phenyl-a-hydroxy acetone .