I have come across something similar. That is, I have a 600W electrical, 2.4kW thermal, ground sourced heat pump with 180metres of pipe buried in our back garden. Initially the pipe just had water in it. After a month of running, it was on the verge of freezing, so I changed to 30% glycol 70% water. This does fix the issue, however the temperature between groundloop and propane is now 3C worse; the glycol is significantly worse at transferring heat.
Some of the discrepancy may be down to the "Partial heat transfer coefficient", pic below found from here:
https://onlinelibrary.wiley.com/doi/...527621583.app2
I've heard 2kW/degC/m^2 before for the temperature drop across a BPHE, but perhaps it's more complex than that, and depends on what the fluids actually are. Water just happens to be excellent, and organics (glycol or ethanol) not so good, hence swapping your BPHE's around so that the larger one is used with the ethanol, smaller one with just water is better?
Loved your vids by the way :-)