This is interesting to me. I read/viewed that experiment with a little alarm. In addition to the aftermarket rims, I believe he also lowered his car with aftermarket springs. In my book he couldn't really test anything since he had no baseline with stock trim. Logically, with no baseline, we can't know whether he got better mileage or worse mileage with those changes. I would suspect worse, particularly with lowering the car.
I believe race drivers lower the body to create suction to hold to the track at high speeds and enhance control. That suction only comes with the expending of energy. Again, can't have much of an idea if the lowered car resulted in the loss of range because there is no base line. If the change of wheel covers can impact the range by ~ 5% then it suggests that other body or suspension changes may have an effect.
A large part of TESLA is the elevation of tech over cosmetics. It seems silly IMO to revert to cosmetics where it works against core engineering and in particular when it might involve the very real and dreaded "range anxiety". It would be nice if cosmetic alterations transparently reported any impact of efficiency.
The other thing I consider is that tinkering with the efficiency of the vehicle as part of an appearance upgrade, is the added continuous burden on efficiency that is paid with every mile driven (if efficiency is lowered) and therefore a persistent associated cost. Maybe I should spend my farkle budget on interior upgrades and leave external sparkles alone unless there is a clear efficiency benefit, again IMO.