The McLaren in 2020 I think was amazing in the wind tunnel but garbage on track. They eventually discovered that their idea of moving the bargeboards forwards, while seemingly logical, worked fantastically in the wind tunnel but on track meant that whenever they turned the wheels to actually go round corners the wake from the from wheels upset the airflow so much the extended bargeboards were useless and did nothing but cause drag and they lost significant downforce. The Wind Tunnel only really gives good information about straight line running with no other cars in front or behind them. Sure they can simulate this too but the limited time they have in the wind tunnel and with CFD time in the regulations mean this is generally not done or done very little.
Indeed and new rules only allowing wind tunnel models of 60% rather than full size, and a max wind speed of 50m/s or roughly 180kmh does not help even working out how a car fully behaves in a straight line, as obviously on track they can do 350kmh plus.
Remember years ago at one team I worked at, budget restraints meant we could only get good time in a tunnel that had a max speed of 50m/s, so about 110mph.
Model looked ok, and even scaling data suggested higher speeds on track should not cause concern.
But first on track test of full car, found massive understeer in high speed corners.
Further investigation in another wind tunnel that cost five times as much to hire, but generared much higher wind speeds, showed front wing was basically generating lift at over 130mph.
Unlikely a situation exactly like that will occur this year in F1, but limited time, on smaller models, and low speed fans, might possibly throw a few curve balls into the mix, when trying to consolidate data onto the full size cars once they get on track.
