KEEP IT SIMPLE....? I am guilty of rambling on quite a bit, though it seems best when trying to help others with overheating issues, to try to stick with the basics....?
First off, an apology to GoMoG... re the height of the expansion tank.... I suspect it may not be not critical in terms of an inch or two, either up or down, as it seems part of the expansion system design requires pressurised air to return the displaced coolant to the rad..? That being accepted...it seems logical that IF the level of the airspace in the expansion tank was mounted below the filler neck of the rad.... then I suspect there is a chance that during the transfer disturbance of hot coolant from the rad to the expansion tank, there may be a chance that air from the expansion tank might just find a way back into the engine.... air ever seeking to rise above water...?
That the heater matrix tends to be located higher than most components in the system, it would seem ideal if the lowest level of the airspace within the expansion tank was kept above the matrix and the pipework connected to it, then that would seem ideal...? I note posts on locating a manually operated air vent located on the pipework feeding the heater matrix, the route of which seems to be purposely extended to create the highest point in the system, which seems a simple enough solution for bleeding air..
As it seems to date that no one has discounted my thinking on the initial design process of the system being that air pressure created by the engine displacing coolant from the rad and into the expansion tank as the engine works towards normal operating temperature, and that as the engine cools when shut down that the process acts in reverse.... Then perhaps using no more than simple logic, I may have understood that part of the system design...? As ever I am happy to wait and be corrected as part of my educational process..