Not sure it was debunked, maybe a bit deflated though. The replication only found the effect to have half of the power as the original.

Edit: nvm looks like it's garbage lol https://anderson-review.ucla.edu/new-study-disavows-marshmal...

The replication problems of the marshmallow test are certainly interesting, but I'm not sure how they are relevant to the point being made by Robert Sapolsky in the quoted passage.

His point has nothing to do with the predictive powers of the test. Instead it's that the children who succeeded in deferring gratification succeeded by means of effective strategies rather than by stoicism or genetics.