I think you've misunderstood a couple of things. Or more spefically: I haven't been clear enough on a couple of things.
I know I can think on a PhD level, because I - was smarter in some eareas than my teacher in cs (who had a PhD in cs) - I've worked with people who mastered cs and during that time I felt like I was thinking on the same level as they were. - I could try to do an IQ test but most (smart)people notice whitin a matter of minutes if the person in front of them is equally smart.
The thing you're missing tho, is the following: I'm Mike Ross before he meets Harvey; I can pass the LSAT of cs but can't be a competitive programmer yet.
If i <em>was</em>, I would apply as an engineer at Google and learn these things in my 20% time.
The Harvey I'm looking for must be able to teach me ao things how to be a good one, and at the same time challange me to bring on new clients etc etc.
Eligibility for PhDs isn't calculated by pairwise comparison.
It's judged by original research. Which requires a lot of study, a lot of study, just to take you to the outer boundaries of the sub-sub-sub-sub-field your PhD happens to be in.
One subject I was taught in a quasi-postgraduate setting (in Australia we have a kind of mini-PhD called "Honours") was evolutionary systems. In one subfield of that topic, the original paper appeared about 30 years ago. A year later there were a few dozen papers. Within five years there were tens of thousands of papers. Conferences. Study groups in computer science departments around the world.
To get a PhD in that field, you need to exhaustively search the literature -- surveying and reading potentially hundreds or even thousands of papers and projects -- just to figure out what has yet to be researched. Just to get to the starting line. That takes several years.
PhDs are hard work.
My advice, go to college for a CS degree. Many companies won't hire someone who doesn't have a degree even if they have "real-world experience". In addition a CS degree will give you a good, solid foundation on which to build your skills. Some of the things you learn in a CS degree aren't well documented elsewhere and are difficult to learn otherwise. If you really are as smart as you think you are then you can take a very heavy course-load to complete your degree faster.
How do you know this? More pertinently, how do I know this? How can you convince others that you are "at PhD level"?
Well, one way to do this is to do something "at PhD level". Build something which (i) is appealing to the kind of people whom you would like to have as colleagues, and (ii) shows your mastery of the full stack (as you define it). Put it up on GitHub, and let others see it. Show them the code, as the saying goes.
What would such a thing be? Knowing what the "interesting" problems are is part of "being at PhD level". So you probably already know what to solve, or it will come to you if you think about it a bit.
Good luck with your efforts!