The problem here is that you're segregating the "world's best programmers" without explaining how you selected them. The more relevant question here is what makes someone the "best" at something.
You're also using a limited angle of success (Larry and Sergey's business success via google). Many of the other "best programmers" on your 1000 person list may have non-commercial success (their creative project leads to a highly cited paper for example).
If you just took the top 1000 programmers (again assuming you could) in 2003, it's unlikely that Mark Zuckerberg would have been one of them. Not that he wasn't a great hacker, but what metric could have demonstrated his creative/business abilities as an undergraduate?
Of course luck is involved in anything. Einstein was "lucky" that he happened to be obsessed with the relative speed of light to different parties at a time when we knew a bit about how light/gravity etc worked. Perhaps if he'd been born in Newton's time, his style of intellect would not have had the same impact. Nonetheless we can, in hindsight, only measure someone's achievement, not his/her luck.
No doubt hyper-achievement (like Einsteins' or Zuckerbergs') is evidence that the achiever possesses a high degree of work ethic AND talent with at least a bit luck thrown in. No doubt that you would agree Hyper-achievers are a subset of "Potential Hyper-achievers" (those with talent & work ethic) and someone's level of achievement will roughly correspond to his/her personal characteristics.
In a particular poker game, it's mostly luck whether you come out ahead, but over time (as in life) those who practice harder and play smarter will (usually) come out on top.