I personally find it pretty difficult to make a comparison of the H50 to different CPU blocks. The CPU temps that would be measured in review are also heavily dependent on the radiator and reasonably dependent on the pump.
You could compare blocks connected to a pump and rad that are similar to the H50, but that's still kind of obtuse, because in custom loops hardly anyone would ever use a rad and a pump fit for an H50, since they have the ability to choose others that are bigger and better by far.
So then if you compare the H50 to a typical WC loop (and "typical" is by no means easily defined) then you're comparing apples and oranges, because the CPU temps will be reflecting differences in more than the CPU block.
I mean, I know you're trying to justify, but making this comparison seems to be impossible or irrelevant in my mind, however it is done.