As you said, the difference between 62 and 57% five year survival is not much, but especially when you realize that often the 5% difference will die in year 6.

Rather than a % to five years (which is a widely used but I think a flawed metric) How about an average time until the end of some kind of quality of life? Or time until you check into a hospital for the last time and not include the time they can keep you alive? If that same data (62 vs 57%) were expressed as average survival time the difference might be 6.5 vs 6 years. That is something that I think is clearer to people.

In the US we famously spend something like 1/3 of all health care dollars in the last couple of months of life. If we pour a lot of money at a terminal cancer patient to add a few months of clinging to life then our 5 year number will be higher than elsewhere but is it money well spent if the patient is bedridden and in pain managed by opioids? If that 62 vs 57% number reflects no real difference in care or outcomes but only a bedridden few months then I would take the British system.

As usual, it's impossible to distill a very complex issue down to one number.