I hate the spin given to that experiment because to me while the results might be predictive of later performance I think they are as much an indication of the child's level of trust in random adults -- in order to wait you have to believe that an unfamiliar adult will indeed keep his/her promise -- than of any innate quality. So are we selecting a trait (ability to defer pleasure) or are we selecting for healthy relationships with trustful adults in early life? Or a mix of both?

You will notice that popular wisdom ("the bird in the hand...) runs against the experiment's preferred outcome.