Surely it's meaningless to compare the efficiency of slavery vs other systems, since your set of resources is completely different.

You could if you look at e.g. the crop yield (ceteris paribus). I don't know why you would, because what sane conclusion could you draw from it?