Well, no, if you have a preponderance of evidence supporting a position, effectively no single "existence proof" data point would really convince large numbers

Well, no, if you have a preponderance of evidence supporting a position, effectively no single "existence proof" data point would really convince large numbers of scientists to change their status quo belief. I would love if you could give some examples. One that sort of seems like it could qualify is the discovery that gastric disease is mostly caused by a bacterial infection, rather than stress, but it doesn't really seem to qualify.