Bill Connelly who does SP+ for ESPN had a tweet pre-Covid that I cannot find that Illinois was something like 85th in expected turnovers and ended up like 8th in actual turnovers.
i do not believe that type of expected vs actual turnover discrepancy is sustainable.
if you are a believer in Connelly's Five Factors he says that turnovers are by far the "squishiest" and hardest to duplicate consistently. So if we are relying on turnovers, and it becomes a "thing" it will have more variance than explosiveness, efficiency, finishing drives, and field position.
you can read Connelly explain his take on the five factors here: