your talking about a NaBH4 reduction of an unsubstituted (no subs on the ring) imine.
The problem here is imines aren't as stable using unsubstitued phenylacetones.
Huh? Do you have any references whatsoever for this difference? I could understand if there would be such a difference with an aromatic imine (i.e. from a phenyl-1-propanone), but not with a phenyl-2-propanone.
And as you can see in https://www.thevespiary.org/rhodium/Rhodium/chemistry/redamin.aqueous.nabh4.html (https://www.thevespiary.org/rhodium/Rhodium/chemistry/redamin.aqueous.nabh4.html)
aqueous NaBH4 reductive amination can be performed in almost 80% yield with P2P's having both electron-withdrawing (fluorine) and electron-donating (methoxy) substituents.