There are good and bad code schools, but knowing about compiler design and binary trees is probably not going to help you that much with your web monkey job. I think you and many others drastically overestimate the value of the CS degree, and you certainly don't appreciate the difficulty of being able to learn several complex web stacks in a few months starting from zero. The pace set is brutal if you already know the languages.
CS grads should ideally know nothing about the practice of programming. Dijkstra never wrote any code except with a fountain pen. A CS course should give you his skillset: being able to analyze programs in an academic sense. This is assumed to be representative of job performance by people who have no other measure. At the very least it should not be expected that it takes less time for the CS grad to pick up the bootcamper's skillset than it took for that person to acquire it. Most likely it will take far longer unless you're focusing on resume-driven development. The reality is that CS courses aren't likely to give you much of an edge in a coding bootcamp, let alone the workplace. In the long run there may be an advantage to the CS degree if the bootcamper can't pick up the relevant bits on the job, but otherwise the real difference is that you can expect the bootcamp grad to be productive immediately.
I strongly disagree with "CS grads should ideally know nothing about the practice of programming." and it does not even describe real life CS grads. Not knowing theory is a problem, knowing also practice is not a problem.
If you was able to learn compilers and implemented one ( as most of those courses require) then you not just have aptitude, but already can read code, can debug, learned to split code to smaller chunks developed etc. In practice, all CS already encountered databases, sql, at least one Web framework, did projects etc.
Web stacks are not that complex and complex parts repeat between frameworks. They seem complex if they are first thing you learn.
Condescending, incoherent, and wrong. If you think that CS is about how computers work in practice you should take that up with Dijkstra and Knuth. It's an absurd position. If you think that 'can read code, can debug' is the benefit of a university education then what about that takes four years to teach?
I'm a self-taught programmer. I talked my way into my first coding job with no experience and learned as I went. For the theory side of things I went with MIT and Stanford's online coursework, and the relevant papers when appropriate (e.g Turing 1936, Shannon 1948, McCarthy 1960). These days I'm getting into Haskell and J -- I hate being outgolfed by the array languages. Recently I audited a couple courses for a local coding academy. At the coding school, the default expectation is that you will be working with a framework that you have no prior knowledge of, possibly in an entirely new language, and expected to complete three to five projects per day, with specifications, unit and integration tests, and a meaningful commit history, five days per week for twenty-odd weeks. Most people require significant amounts of study time outside of class. This is not descriptive of CS coursework. It's almost like there are two related-but-distinct bodies of knowledge are being pursued in different ways!
If you have the idea that bootcamps are easy then I definitely invite you to try one sometime. Discard the idea that what the university and bootcamp are offering are somehow equal. A CS education may or may not be a better preparation for a career in programming, but it's not good job training, and that's on purpose. If things were otherwise, coding schools would not need to exist, would they?
A CS degree is a loosely-correlated proxy for being able to code. It happens to be very easily measured. This leads small minds to believe that it is a good proxy, especially if they happen to have CS degrees. I'm sure the rationalizations and condescension come naturally.
CS grads should ideally know nothing about the practice of programming. Dijkstra never wrote any code except with a fountain pen. A CS course should give you his skillset: being able to analyze programs in an academic sense. This is assumed to be representative of job performance by people who have no other measure. At the very least it should not be expected that it takes less time for the CS grad to pick up the bootcamper's skillset than it took for that person to acquire it. Most likely it will take far longer unless you're focusing on resume-driven development. The reality is that CS courses aren't likely to give you much of an edge in a coding bootcamp, let alone the workplace. In the long run there may be an advantage to the CS degree if the bootcamper can't pick up the relevant bits on the job, but otherwise the real difference is that you can expect the bootcamp grad to be productive immediately.