Nothing in this article proves anything about whether you need to go to college. If anything, it indicates that Peter Thiel may have picked the wrong kids for his experiment.
While the article certainly lacks hard proof, the "he didn't pick the right kids" argument is even worse. The ratio of successful college grads to non-grads may not show causality, but at least it is some sort of metric. Until someone can quantify "the right kids", it beats the alternative.