I'm not sure I understand your point - are you arguing that finding cancer earlier does not increases chances of treatment and survival? Do you have any reliable data on that?
The conclusion: "Although 5-year survival is a valid measure for comparing cancer therapies in a randomized trial, our analysis shows that changes in 5-year survival over time bear little relationship to changes in cancer mortality. Instead, they appear primarily related to changing patterns of diagnosis."