It's really hard to generalize.
I think "person who is a top 10% but not top 5 Stanford undergrad CS senior" who then goes to found a YC-backed startup which ultimately fails after 2y is ultimately better off than the person who takes the "good dev job" at Google or Facebook. When the founder is looking for a job 2 years later, if it's via acquihire, it is probably a wash on cash (due to taxes...potentially quite ahead); has almost certainly had more public visibility and thus potentially is in a "bidding" situation for his talent, etc. I doubt the 90th percentile Stanford CS grad gets more than $150k cash, $250k total compensation, in year 1, and probably not more than 200k/300k in year 2, at a tech company.
A person at the 90th percentile, all things being equal, probably isn't good enough to get one of the $250-500k hedge fund programming jobs.
Taking the good dev job/employee #1 at the hot startup which happens to do well is better than being founder of the failed startup.
Obviously, the person who starts Facebook is ahead of everyone.
Partially it's that equity compensation (in an acquihire/earn out) is tax privileged, part is that it's essentially forced savings. From $100k to $200k, you lose a lot of your income to basically bullshit -- US/CA taxes and somewhat higher living standard, higher student loan repayment, etc. -- where if you can essentially bet that "pre-tax" on a startup, you probably come out ahead.
I'm not sure how this applies to someone in the bottom 90% of Stanford, or in the bottom 99% of the world. Probably the best bet is to somehow get into a top startup in a role where performance isn't so critical to the ultimate success of the company, usually after Series B, or at a big company.
Someone who is the top CS grad for the year, or who is later in career with exceptional talent, or who has a burning drive for a specific area (e.g. Steve Mann in wearables), has an entirely different analysis, too. At that point what actually matters is role.
I personally am happy doing a startup to accomplish my specific change-the-world goals and would be equally happy doing so at a large company; it's just that it would be much more difficult in ways I don't enjoy at most large companies (budget and multiple hats at a startup; stupid politics at a big company). (If I had more of an aero/engineering background, I'd probably prioritize space, and try to work for SpaceX; as it is, the only things I'd be qualified for there are IT, and their IT is windows shit, and not core to the success of the enterprise.)