Define "success rate," because there's successful people out there from all walks of life. Larry and Sergey are clear examples of the academic route. Steve Jobs is not.
I think you're trying to put things in nice little buckets when the truth is that the world's a messy place. Also, it feels like you'll suffer from confirmation bias, as well. Maybe you're just thinking about the latest single-page-web-app startup. What about places like Bose, which came out of MIT? Or really any of the myriad companies that have been born out of academia.
Rate would be tech startup successes per student. Of course I agree that the rate for academic environments would be greater than zero, and some great companies too.
You're already comparing unlike samples. The base rate for startup success is constrained by ... starting a startup. Folk who rock up to the YC offices are already self-selecting into that group.
I think a better measure would be wealth net wealth at particular intervals -- 5 years, 10 years, 25 years, 50 years.
I think you're trying to put things in nice little buckets when the truth is that the world's a messy place. Also, it feels like you'll suffer from confirmation bias, as well. Maybe you're just thinking about the latest single-page-web-app startup. What about places like Bose, which came out of MIT? Or really any of the myriad companies that have been born out of academia.