Debunking the Shock Cooling Myth

We are running 2 Lycoming 235 Pawnees (one wide deck & one narrow deck) and a Continental L-19. We went through a spate of cracked cylinders (3 - In two different aircraft as I recall) about 4 years ago but I am not sure it had anything really to do with cool down procedures. Then the problem went away with no good explanation. Our method is to hold 2000 - 2100 rpm for a minute after glider release while descending then make a normal landing. I suspect that there are other issues involved in cylinder cracking: quality control in cylinder manufacture, torque variations in cylinder installation, localized vibration caused by cylinder/piston/ring fit, front vs rear cylinders (they cool differently), valve adjustment, and plain bad luck.
ROY - GBSC Boston (where it is really cold this morning!)