I’d totally bet this down to 20% if it were a real world resolution criterion for 2050

self-resolution criteria make it nothing to do with prediction

This question is a cure to the never-resolve cancer that takes up a lot of Manifold O2

It is very unlikely there will be a single treatment since cancer is not a single disease, rather a very varied group of diseases. I presume the resolution does not require a single cure for all types of cancer, right? A single cure is very unlikely, curing subtypes separately is IMHO somewhat plausible(well unless you count something like "shoot everyone diagnosed with cancer" as a resolution- that would indeed reduce deaths by cancer to 0 and is already "available" to some.)

Is there a potential medical reason why this could not be done?

@LightLawliet A possible problem would be requiring exactly zero mortality. Even for diseases we would generally consider to have a "cure" (e.g. scurvy, streptococcal tonsillitis), the mortality with correct treatment is certainly low, but not quite zero (e.g. because for people that are already frail, the disease might have done irreparable damage already before it is diagnosed)

resolve to market % is a better way to do this sort of question