Your view of avoiding a career in a large business organization, I believe, is fully correct: Starting at age 22, to age 70, assume 70 will be the Social Security retirement age, you will have 48 years of work, and the chances of doing all that just in big corporations are slim to next to none. One reason is that, over the past 100 years, for a single corporation to be stable for 48 years was rare. Another reason is, the strong but unwritten rule is that the subordinate must be younger than the supervisor: So, by age 45 or so you are qualified only for CEO. So, about the only hope will be CEO of a small organization, most likely yours. So, you will have to do startups anyway and should get on with it.
E.g., assume you start at age 22 and (A) join a large corporation and get fired at age 40 or (B) start a grass mowing service and grow it and expand to higher end clients and some landscape architecture. At age 40 (B) stands to be much better than (A). If being an electrician is better than grass mowing, then good. If computing is still better than being an electrician, then still better.
A career with a license, e.g., law, medicine, electrician, can be good, but you are not considering any of those.
State government jobs don't pay very well, and government jobs are much less stable than is commonly assumed, are really not stable enough to bet your career on. That you have set aside careers with academic tenure or licenses, basically you need to be running your own business. Period.
Millions of US families are supported by men who own their own Main Street businesses. If computing is a better opportunity than, say, HVAC, auto body repair, or an independent insurance agency, then so much the better, but you still need to run your own business.
A BS from Harvard, Princeton, Yale, or Stanford can bring a lot of prestige and let you meet some people who are or will become important, but running your own successful startup will make you a still MORE important person. Uh, a Harvard guy seems important because of his chances of running a successful business and the people he met that might help him run a successful business: If you are already running a successful business, then basically you are ahead; the Harvard guy may want you to hire him!
Yes, there is a joke: "If I need a Ph.D., then I'll hire one.".
So, if you don't want prestige, then you are free to go to a state school and save a lot of money.
Still, college will cost you about four years of time away from your business (likely a big problem), four years of income from your business, and enough cash out of pocket to buy a house or nearly so. Big question if college is worth it.
If you don't go to college, then work hard to pick up some of what you missed by some high quality directed reading or a few courses as a special student.
So, you need to learn some things.
===People===
For your life both inside and outside business, you need to learn some things about people.
For business, you need the basics of organizational behavior -- even at best people are goofy to work with.
More generally the fraction of badly mixed up people in society is surprisingly high, and basically you need to be able to detect, maybe diagnose, and likely avoid these people on your own, quickly.
One rule is, if talking to them leaves you somewhat or more shaken, then you just detected some bad intensity and struggles within them; stay away from them.
People can make such detections at an early age. Indeed, in:
T. Berry Brazelton, M.D., What Every Baby Knows, ISBN 0-345-34455-3, Ballantine Books, New York, 1987.
is the flat statement that when a child is rejected by his peers it is because the peers detect some intensity that scares them.
For more, there are four neurotic (now just an informal word) styles: paranoid, hysterical, obsessive-compulsive, and psychopathic passive. You need to understand these; once you understand, you will see examples fairly frequently. Then a lot of what they do will be understandable and predictable. A good start is just:
David Shapiro, Neurotic Styles, ISBN 0-465-09502-X, Basic Books, New York, 1965.
The author has an MD.
More generally for people 101 read:
Erich Fromm, The Art of Loving, Harper and Row, New York, ISBN 0-06-080291-X, 1974.
Fromm was a Freud student.
It's a classic, originally published in about 1947.
Men and women, whatever else they are, are not the same. For some insight into women, read:
Deborah Tannen, You Just Don't Understand: Women and Men in Conversation, William Morrow and Company, New York, ISBN 0-688-07822-2, 1990.
Here Tannen wrote a popular book about her academic specialty; she has been a professor at Georgetown.
I am NOT an expert in psychology. If you can find a college professor who is good a expert on people and teaching such material well, then at least get their reading list and maybe take their course, if only as a special student.
===Science===
Biology, chemistry, and physics stand to be of increasing importance; you might want to know more about each of these.
Just in physics, just for computing, you should be able to do basic energy calculations. Also in electronics the practical consequences of Maxwell's equations are good to know. Quantum mechanics stands to become of more importance in computing.
===Business===
You might get some benefit from accounting, finance, and labor law.
===Mathematics===
Math is the most precise knowledge we have, and knowing some of it can help provide a benchmark we can try to approach outside of math.
Also my opinion is that math is the most important field, more important than computer science, for the future of computing, especially in entrepreneurship where it is good to have some powerful secret sauce that delivers valuable solutions and is difficult to duplicate or equal.
I have some expertise in applied math -- a good Ph.D., a lot of successful applications, quite a lot of college and graduate school teaching, and some published papers.
===English===
Good skills in critical reading and in writing for a purpose could be crucial for you.
===Computer Science===
I am not convinced that college courses in computer science will do much for you in the future of computing.
So, we have subjects:
People
Science
Business
Mathematics
English
One way or another, you should seek to make good progress with each of these topics and maybe some more. For nearly everyone, college is the best approach.
Still, you have learned practical computing on your own. And in the rest of your career you will have to continue to learn a lot, more than people learn in college, on your own. So, it is not clear that you can't learn on your own what others learn in college.
One approach would be to compromise: Get some college course catalogs and see what the main courses are. Go to some college campuses and see what texts and other materials the courses are using. In time it will become clear what the most important texts are. Then for each such text, scan it, maybe just in a library. If a text or course looks really good, consider buying the text and/or taking the course, if only as a special student.
If as a special student you see that you actually want a degree, then go for it then.
In the end in life, you have to get the work done; you can teach yourself things that will help you and you have; college will teach you things, too; but you still have to get the work done, and college won't do that for you. Indeed, if you go to a good graduate school, then the main messages will be (A) learn the basic material for the qualifying exams on your own or nearly so and (B) do some research, mostly on your own. In the end, college or not, you have to do work of your own largely on your own. You got started early; good for you.