This question is about mathematics only. My forethought is that courses are usually based on textbooks, but they have less content than the textbooks. There are less problem sets, and chapters are skipped. Is it true that you can learn more by self-study with textbooks rather than college courses? Given that one could find help in the internet.
Without a constraint of time...
If you mean "because self-study can be done over a longer period than a typical college quarter or semester", then this is as relevant as the other points you make. You don't elaborate on it in the body of your question, but a time-constraint can work both ways -- not having one can be great, because you can work at your own pace and not miss things along the way. [I actually consider the typical term length to be a weapon of math destruction, particularly because students who can't assimilate concepts at the prescribed rate often find themselves in the fail-retake spiral, never completing their degree.] However, working with a time-constraint can be good, particularly if it builds motivation to keep at it. I know I could learn to paint on my own, but having an instructor give me deadlines really keeps me working.
If you mean "let's not consider the constraints of time in this discussion", assuming person A takes the course face-to-face and person B self-studies their way through the material, then I would bet that person A learns more, as I'll explain below.
...courses are usually based on textbooks, but they have less content than the textbooks. There are less problem sets, and chapters are skipped.
Mainstream commercial textbooks try to appeal to a wide audience, so they often include more information than any course requires or has time for. So in this sense, you're right -- a student who self-studies could certainly take on more information by just finishing the book. But wouldn't this also apply to students in a face-to-face course? Every few terms, I have a student who just decides to work through that last chapter because she's interested in the subject, but this outside of the norm by far. My experience, and research I've read, says that students don't read textbooks anyway, except when mimicking a worked example for a homework problem. This may be one downfall of face-to-face classes where instructors explain every last thing in detail -- students are conditioned to not read the text because nothing bad ever happens when they don't, just as people in general are conditioned to not read End User Licensing Agreements.
Given that one could find help in the internet.
If you're an independent learner who knows how to search for content, and you plan to work through a book at a matching level of difficulty, and you set aside time to actually do it, then you could definitely cobble together your own course and learn more than is required of you in a typical college course. You will find enough help on the internet to address nearly every potential issue that could come up with misunderstanding the material. However, a procedure for diagnosing and treating this misunderstanding will be costly time-wise, either for the learner or for the person they found to help them online. One thing a teacher can do better than a machine is to ask leading questions to determine what a student really knows, and then build that toward learning a new thing. If you're on your own and you really misunderstand something, it may be quite difficult to figure out the root of the problem by searching online. Getting on a discussion forum (such as the Mathematics stack exchange) could be the fastest way to get this feedback, but the answers there can vary so widely in their scope that figuring out which one to focus on will be a new challenge, every time you ask a question.
In short, I would say "yes" to your question, in theory. However, I don't see students reading the book on their own, even when they're in my classes. I think having a deadline in an actual course provides motivation for students to actually keep working on the material, and having a human teacher with whom to interact will provide trained, tailored feedback to address misunderstandings and prompting. Finally, a community of other students is a great environment for learning. Having/getting to explain what you're thinking to someone next to you helps solidify your own ideas. This may never happen with self-study.
It is a mistake to think that one learning method is best for all students. There may be some students who learn more by self-study. There may be other students who learn more when taking a structured course with the help of an instructor.
Some students, if studying on their own, will soon quit. But a regular lecture, with homework and deadlines and the threat of a failing grade may be enough to keep them going.
In a similar way, some people hire fitness coaches to force them to do their exercises, since when they are left on their own they don't do it. Or don't do enough of it.
This is a hard comment to answer given the caveat about without a constraint on time. (If anything, one of the advantages of self study is more time efficiency versus lectures, as well as more flexibility to move faster.)
But on the uncaveated question, I think yes, you can do as well or better with self study. In particular, with lower level courses that have settled content and well groomed texts, full of exercises (and answers).
I think the downsides of lack of forcing discipline and social interaction are much more serious than the issue of not having the instructor to ask questions of. Provided you pick a helpful text (ideally a programmed text, but certainly not a "derivation left to the reader" text), than you can get almost everything you need, especially if already a strong student. (I would not recommend such self study for a weak student.)
If you take a disciplined approach of writing down questions in your notebook (as a sort of "parking lot" to allow you to proceed), you will often find that with the benefit of time (just later in the chapter or homeworks), you are able to answer your own question. In the few cases, you can't, the questions are easy to take to a Q&A site (again, I recommend trying to parking lot and then ask versus stopping). Note, this is also the best way to use your instructor!
Is it possible, that self study will leave gaps? Sure. But then again, how many instructed students are gap-less? Let's be real about the comparable. For instance, if I had to choose between reading the book and working extensive drill problems versus listening to lectures and doing minimal drill, it's a no brainer, which approach is better. See also the following essay on how lecture is overrated:
"I like watching lecture videos, and I’d watch them whenever they were available. Videos are nice, and watching a video is certainly less strenuous than reading a textbook.
But after having finished the classes, I came to an unfortunate conclusion—the videos don’t matter too much. Having a video explanation of a concept is nice, but it’s rarely superior to the same explanation in text. Text even has the advantages of searchability and nonlinearity, features missing in video.
What mattered was having practice problems and projects. Under time constraints, speeding up lecture watching to spend more time on practice or insight-building tactics was almost always a good tradeoff. Knowing I’d need to pass the exam later, I’d always pick a course without lectures than one without practice problems."
(Now of course, you can make the comment that you could have BOTH lectures and drill, offline. But please consider how the conjoint either/or thought experiment reveals something about the value of instruction. And especially if we consider time efficiency.)
I also quite like the comment about strength training as an analogy. Consider that Lou Ferrigno worked out for most of the beginning of his career in a little home gym. Diligent study is several times more important than having an instructor. (Sorry, people-who-get-paid-to-teach.)