Maybe that kid who made the wrong website put more emphasis on things that places like MIT prioritize instead of reading about "useless" subjects like learning how to create secure websites? The issue is comparability. You can compare by SAT scores and grades far better than by a metric like "built a secure website". If places like MIT prioritized such things more, they'd become even more unfair. Don't get me wrong, I'd much rather work with you on a team than with someone who cares only about their personal success above the team's success, and can't listen to input as to why their approach was wrong.

As for the Zuckerberg example, I think it's a prime example because a social network website is comparatively cheap to create, but the key question is to get people to sign up. By launching first for top colleges, people who want to stay in touch with their "successful" mates would want access as well. Someone without access to a top tier institution's social life would not have been able to pull it off.