If correct, it throws the fabric first/insulate, insulate, insulate heating cost/emissions reduction mantra into doubt doesn't it?

re trenches:
We dug our own trench, put in a HDPE pipe and a (diy) heatpump. The trencher looked like it weighed the best part of a ton, had a 900mm long 100mm wide chainsaw type thing at the front that lowered into the ground. It was brilliantly easy to use even for a novice, just walk backwards, it did the work. We cut around 200m of trenches, using from memory 10litres of unleaded. Getting the trencher to us probably took as much energy again. We used 250m of 25mm dia HDPE pipe.
My noddy internet search calcs indicate CO2e of the pipe is 150kg, transporting and digging the hole was perhaps 50kg CO2e. There's bound to be loads more little bits adding up - I think mostly to do with transportation. Our system is tiny, clearly a larger system will have a bigger footprint. For reference, we used 4500kWh/year gas in 2021 - equivalent to 900kg CO2e, and it's now 1400kWh/year elec for the gshp.

My own observation of COP versus power throughput is that as the power level increases, it becomes more worthwhile having a more complex system. Our gshp uses a cheapie but reliable capilliary tube - better efficiency would be found with mechanical or even electrical pressure drop techniques. Our gshp has cheap fixed speed compressor and pumps (600W compressor, 40W pump outside, 20W pump inside), variable speed compressors and pumps are less reliable but save energy. I appreciate saving energy is important, but one solar panel would generate more energy/year than I could save with all the gshp improvements I could do, and PV is so simple to add.