It almost seemed a silly question to ask because I can't understand why Mitsu would stick the 'Ambient thermistor' (TH7) in a place where it is so exposed to sunlight, although they do make the point that this thermistor works in conjunction with the '2 phase pipe thermister' (TH6) which is inside the case.

Perhaps they assume that the whole cabinet is usually mounted quite close to a wall of the house, whereas mine is about 3m away.

At 11:30 it was reading 21c at an ambient from my Vantage View of 15C. As an experiment, I put a piece of plastic trunking over the thermister and now get 17C against 16C.

It shows that you shouldn't ignore the basics!

This positioning of this thermocouple doesn't seem very sensible as it is fully exposed to sunlight, so I wonder if it is providing the most useful data for weather compensation.

Is there a case for providing some kind of shade over the thermocouple to give ambient rather than surface temperature?

If the sensor which feeds the WC algorithm is in  sunlight, it shouldn't be, as @davidb says.  The idea is to measure air temperature.  Weather stations are in Stephenson's screens for a reason and yes a bit of shade (provided you dont create a 'dead spot') would be a good idea.  It may not make a major difference however, other than on cold very sunny days.

This is from my MelPump data and the white line shows the outdoor temperature significantly exceeding the indoor temp of 20C, with the last reading at 24C.

My Vantage Vue says that the ambient temperature is only 17C, but if I point my infrared thermometer at the thermocouple mounted on the outside of the Mitsu case, it does read 24C. (The  peak of around 23C at 10:00am was when the ambient was around 10.5C)

This positioning of this thermocouple doesn't seem very sensible as it is fully exposed to sunlight, so I wonder if it is providing the most useful data for weather compensation.

Is there a case for providing some kind of shade over the thermocouple to give ambient rather than surface temperature?