Even a fairly recent copy of the Moroso catalog perpetuates the "too fast to cool" theory. Since any type of heat transfer improves with greater temp difference, pausing to let coolant "warm up" has to degrade heat flow, and thus cooling. Combine that with the stagnant liquid boundary layer stuck to the hot cylinder head through which only water's crappy thermal conductivity can work, and greater speed will again only help by scrubbing the boundary layer thinner. Look at the cooling passages in modern high performance engines. Thin for high velocity over hot surfaces because the coolant in the middle of a large channel is just going along for the ride.
Or, eat that ice cream cone while driving with the window open and see whether it warms up and melts quicker than in still air.
Well then, are the folks who believe they have had even good running engines overheat when running completely 'stat-less liars or fools?
Probably not. The race shops that advocate thermostat removal add a restrictor plate at the same location. This would have the effect of creating pressure drop as the coolant leaves the block, which does 2 good things. First it reduces pump flow a bit, and that improves the relative pump inlet NPSH condition making cavitation less likely. Second, it raises the pressure in the block a bit (which is already over the pressure set by the pressure cap), reducing the chance of local boiling at hot spots in the cylinder heads, which WOULD clobber the engine cooling in exactly the areas that need it the most.
