kind of pointless to 2nd guess the designers and physicists Denny, but part of the flaw is in not taking account of the variation of the coefficient of thermal expansion with temperature & pressure, the 5-10* drop in temp across the rad, or the volume changes from pump agitation, as well as assuming the tank is right full when hot .... as you say, let it find its own level, it will
always be well below "full" [cold] but above the tubes when the system is working properly, the mistake folks make is filling it right full when cold
and some related trivia: when pasteurizing juice in my large gas-fired kettle, I have to fill to a lower level when the ambient barometric pressure is rising or it will foam over ... a demonstration of pressure affecting expansion of a heated liquid
Bill
