It's hard for me to wrap my mind around your problem as well. You want to make a supersaturated solution of those particular transition metal oxides in liquid Na3AlF6. That's a little outside my experience. But let me try to think of it analogously with aqueous crystal growing.
Say I'm growing crystals of aqueous soluble salts, but instead of evaporating the water, I decompose it instead. Well, that's plausible, but if I add something to the mix, aren't I contaminating the mix with that reagent, and with potential products?
Interesting idea 'tho. Say I wanted pure crystals of an aqueous solute. Suppose I electrolyse the water away instead of evaporating. Hydrogen and oxygen fly away, and the solute crystallizes out. You can't do that -- or can you? If you electrolytically decompose the cryolte, even if aluminum metal precipitates at an electrode, won't the crystals grow? Does that happen in commercial aluminum production?