The rationale behind my thinking is the significant difference in observed mileage with 50% of my grille blocked off in cold weather. The return thermostat would control the amount of coolant from the radiator with the purpose of allowing only warm coolant to enter the engine.
In hot temperatures we get better mileage. It seems from my observations that controlling the heat losses from the radiator to maintain return coolant temperatures regardless of ambient temperature would go a long way towards eliminating the vast difference between cold weather and hot weather mileage.
Water pump bypasses would be a partial solution, but you still have very cold coolant entering the engine when ambient temperatures are low.
My thinking is to eliminate the heat losses attributable to reheating the very cold coolant entering the engine in winter. I have not tried blocking off more than 50% of my radiator, don't want to risk anything engine wise.
As I said earlier in the old days they closed the louvers to accomplish the same thing. Its really significant when you look at the big rigs that have almost all the grille covered to keep operating temps higher.
The lower the temps (some of you here have to deal with 40 below) the more significant the mileage improvement would be. It would be interesting to measure the return coolant temps when ambient temps are very low. This would be especially significant when hypermiling is utilized.
Closing off air is the only real way to do it. You won't be able to just put a thermostat in the radiator like that because on a cold day even a 120 degree thermostat would let the engine overheat before it'd open because the radiator temps never came up.