Hope you do not mind but as cooling issues ever seemed to be a hot topic (pun intended) on the various Morgan forums over the years, I thought to add my tuppence/two cents worth.. (-: More than happy to be corrected if I have made any mistakes..
I can well see that there is quite a bit of room for confusion given that the neck on the radiator of my mid 80`s +8 is identical to that on the expansion tank, thus the rad would take a "normal" pressure cap, whereas on my Mog, the pressure cap is fitted to the hard plastic expansion tank, and as already been explained the radiator is fitted with a blank cap, thus the flexible tube from the filler neck of the rad that connects to the bottom outlet of the expansion tank is free to transfer the coolant as it warms and expands, into the expansion tank.
It is the expansion tank that determines the allowable pressure within the system via the pressure cap setting (15lb on my +8) the top tube on the expansion tank is the overflow, which should only ever overflow relative to the pressure build up in the cooling system rising above 15lbs/sq inch..
On my +8 as can be seen in the pics the water level in the sturdy plastic expansion tank can be clearly seen through the material the tank is made of which is perhaps worth noting as the tank it`s self has moulded lettering with arrows indicating WATER LEVEL.... these arrows point to the lower ridge in the tank where it slims down, as if it were shaped to take a strap... If my description causes confusion I will try to grab a pic of the level markings.... On my Mog the level in the tank reaches those arrows from it`s present position only when at normal operating temp, which for my carb fed and standard tune mid 80`s+8 is 90Deg C
If filled to the indicated water level when cold, the excess is seemingly spat out during a trip, and on returning home with the engine up to temp the coolant id found sitting at the water level arrows, and once cooled returns to the level as can be seen in the pics....
I suspect the height of the expansion tank seems unlikely to be critical...? It would seem logical that as the engine heats and the coolant expands up the tube and into the expansion tank the pressure in the expansion tank will increase. As the coolant drops in temperature and contracts, it seems likely that the pressure built up in the tank by compressing the airspace above the coolant level in the tank, will then force the formally expanded and now contracting coolant, back along the tube and into the radiator.
I took a pic of both my rad pressure cap and blanking caps which I hope to provide a ling for below.
The thermostat opening temp is listed in a Rover 3.5 V8 manual as being 82Deg C (180Deg F)
Reluctant to type this.... But my Mog has not ever overheated, even when Mrs Luddite and I got way too hot while touring in a VERY hot Italy some years back.
Another point perhaps worth mentioning is that even rad caps with the same poundage can operate differently, as can be seen if you open the E type link below. I suspect the one with the small valve that is open to atmosphere when there is no pressure in the system, does so in order prevent vacuum build up being created by coolant contracting and thus avoids the rubber coolant hoses collapsing.
Radiator Caps Explained (coolcatcorp.com)
https://www.tm-img.com/image/SvUAnhttps://www.tm-img.com/image/SvBu4
