In the abstract, I agree. One could create a utility function expressing the necessary compensation (for them) in exchange for taking on a certain amount of risk. In the concrete, I'm not so comfortable with this on a population scale. People in general are not good at understanding risk, and they are not good at understanding medicine.
Procedure X has a 10% chance of killing you, a 1% chance of leading to necrotizing fasciitis that will not kill you but will require an amputation, a 3% chance of disfiguring scarring, a 4% chance of in-procedure myocardial infarction that will leave you alive but in heart failure ... etc. If we even had an exhaustive list of the actual risks of the actual procedures in the hands of the actual surgeon doing the procedure (we don't), it would still be impractical to get any particular person to understand the whole array of risks that they face, not just mortal risk.
We would need a much better way of measuring 'understanding' than I believe we have right now for me to get onboard with compensation for extremely high risk procedures for donation. And that's assuming a coldly rational populace.
