Survivorship bias is real, but so is remarkably low return on investment. Its just harder to measure in people.
College is NOT an absolute fast track to employment. But it's an excellent reliable way to do that. Plus it's a fantastic place to find out more about yourself, meeting people, friends and for relationships. A lot of people have the college drinking perception. Which there is too much drinking, but there are a lot more ways to meet people in college where you're not drinking than in the working world.
Again just respect the work you and other people have done to get where we all are. Just to say "most people shouldn't go" is ridiculous. We want MORE people going. Show them how to do it in a financially responsible way. If that's too boring then stick to what you know.
I would even go so far to say that OP's anecdotal quotes from friends in "investment research" weaken his conclusion (i.e. I'm (100%) sure I'd be a millionaire if I'd started my current job 4 years earlier. But in order to get my current job, I must've gone to school for 4 years.)
For example, there's a subtle difference between these two stances...
your title: "A dirty secret: Most people probably shouldn’t go to school"
your conclusion: "So here’s a dirty secret of our times: you MAY just be better off not going to school."
If you're arguing that education could be faster and cheaper, I don't think that comes out in your post due to other distracting claims and anecdotes.
To me that just means folks need to be more thoughtful about the degree they select... after all, a quality, valuable education isn't something you're simply given, it's something you take.
It's unfortunate that this kind of thinking is so pervasive. You see it in a lot of contexts... I've seen this in folks who manage to raise themselves out of poverty and become successful, only to turn around and oppose social programs because, well, they didn't need them!
The author can't be faulted... this type of reasoning, a combination of survivorship bias, hindsight bias, and confirmation bias, is extremely human. We're wired for it. The hard part is realizing when you're doing it so you can try and catch yourself. The clearest remedy is the simplest: use data, not anecdotes.
In this case, the data supports the view that post-secondary education leads to long-term advantages. Of course, that needs to be balanced against the increasing debt load students are burdened with. But the reality is, generally speaking, more and more jobs are requiring a post-secondary degree, and advising people against getting one is generally speaking bad advice... while there are the odd outliers, they are just that, outliers.