It may not be ethical to do this under the guise of administering medical practice even if there is informed consent. Part of a doctors duty is to be the rational actor to say “you could have 6 months to live and this procedure has no evidence that you’ll live any longer, in fact, it could shorten your life to the time of your surgery”.

This is advocating for civil engineers to willingly approve on a bridge they have no evidence won’t collapse, just because the current bridge needs serious replacement.

Define "slightly better shape". Because that's now your inclusion criteria. You have to be sick enough that "Or this might not work and kill you" seems like an acceptable risk, but not as sick as the guy they chose by arbitrary and hidden metrics primarily based on looking at the outcome retrospectively.