Afaik there are no concerns from the protocol since we check everything lies in the subgroup, but there maybe some concern in one corner of the proof. arkworks-rs/algebra#726
We'll likely just ignore this since better ring VRF protocols never touch bandersnatch, but it'll maybe come up again in sister curves.