Well, maybe if you listened in class you would realize the benefit of what they're trying to teach you.
If you really think you can understand complexity theory, algorithms, and advanced data structures in "an afternoon or two", I think about 90% of it went right over your head.
Yes, the traditional courses do teach you that as well - to some extent - but the point is it's important to see how different fields do it. I think my classes in algorithms are as useful to my research in physics as my classes in physics are to my research in algorithms.
Those physics classes will teach you to think about problems abstractly. In my experience, my knowledge of physics has helped me think about algorithmic problems significantly. Even if you don't use the physics itself, it's good to learn.
The accounting might not help you so much for CS, but it could be useful. And a few liberal arts classes are always good for breadth of education, and making you realize that you can always pick up a good book and read it, and that not everything you do has to be so math and science oriented.
Also, for the record, what you would learn at a trade school would be "programming". What they teach at university is "computer science". There is a pretty serious distinction, and I don't really believe you can learn computer science properly without a formal education (even if that comes from self-studying formal books - but teaching yourself javascript is not computer science).