Can you clarify when you say, “phases being quite different in voltage,” you mean that each phase has a different Vrms? Or are the Vrms amplitudes actually the same, it’s just that they are not aligned in phase?
If the latter, you can try my firmware listed here with a single emontx3:
Emontx3-continuous - New continuous monitoring firmware
The calibration constants in cal.h can be set to IPH1=-120 and IPH2=-240 degrees , to rotate two of the phases into alignment with IPH0. As long as Vrms for phases 0,1,2 are equal, it should be fine.
Craig