Maybe you give it to the authors of a few papers, but even then you'll struggle to capture even a fraction of the necessary preconditions.

The successes also rely on observing the failures and the alternative approaches. Do we throw out their credit as well?

The list would be longer than the human genome paper.