Hi Jeremy,
Thanks for this input! Great to get a sense of the relative value of PEPs vs TM65 estimates, all makes good sense and matches my perception when comparing data for the English language PEPs with the TM65 basic and mid-level data (and also clear from the methodology behind TM65 vs ISO 14025). Yes, je pense que mon français n’est pas suffisant…although I know some Francophones who may be able to help with translation as needed. Great note on the NA vs EU electrical protocols and the associated difficulties in functional equivalents - that seems to be a consistent theme in the feedback, I appreciate it!
Part of my research involves identifying preliminary steps towards accounting methodology. There are versions of this in which placeholders are used for MEP equipment, at least as defaults, to hold values for whole building carbon accounting. There are versions in which product-level data is deployed to inform equipment selection; it’s clear there are differences in emissions between different units, and certainly across different system types, but it’s questionable if the data is there for HVAC equipment to conduct accurate product-level comparative analysis. The refrigerant impacts seem pretty reliable to quantify, and are actionable in unit specification - it seems reasonable that one could quantify refrigerant mass, type, and apply an ASHRAE or equiv leakage rate to evaluate the refrigerant emissions difference between a monobloc A2W system with hydronic distribution and a VRF system, however there would likely be a greater level of uncertainty in the embodied carbon impacts between the equipment based on data quality and comparability.
Exciting times in the world of carbon accounting 
Thanks to you all for the great feedback!
Jacob