Scientific, closed testing is impossible in the real world. Temp/pressure/alititude/air density/wind/humidity/driver/traffic conditions/etc etc. Are always changing and for the most part uncontrollable. All we can do is take averages over multiple tanks and compare.
I suppose we could start taking logs if obd2 data, and I might on 2 of my cars, but there is a lot of work there sifting through it.
Recently all my averages have tanked and I suspect winter blend is back, historically all my vehicles take a 7-10% on winter blend which is where they are all sitting now, but using tcq3 seemed to have reduced the negative effects of ethanol on 3 of my vehicles, plus small engines.