I didn't *think* so but now i'll have to test again (tonight). It should be easily reproducible: just send power from a CESU to another CESU, twice, where one route goes via a LV-transformer, and the other route doesn't, and see if all the power reaches destination in both cases.
I should test this also. I'm also not 100% my current strategy isn't wasteful, either. I'll test this:
MFE --> Batbox --> MFE
MFE --> 4x Batbox in parallel
Start with everything empty except source MFE.