i seem to recall a novel by jean paul sartre "l' autodidacte" (the self educated man) in which the title character had learned only by reading. he was sort of a pitiful ignoramus in the opinion of the author it seems.
Of course you are not really self taught, you are a student of the author of your book.
Now sadly, it is normal practice in most universities to use books that are written by mathematicians or even physicists who are actually less well informed on many topics than are the professors teaching the class.
so to take as your teacher the book's author instead of the class lecturer may well mean you are choosing the poorer of two alternatives.
The other problem is that if not truly self taught, you are indeed truly self examined, and that is a major problem. I.e. you are the one deciding whether you know the material or not, or at best you are exposing yourself only to some possibly trivial problems in the book, and maybe to one attempt at a comprehensive exam, to exempt the course.
Now it is very very hard to design an exam whose successful passing truly guarantees that you know a subject.
Of course it is possible that you do really know this material well, but it is not usually the case for most students.
If you are at anything like a decent school, your professors are active researchers, whereas most textbook writers are not so current, but are writing things that they themselves learned long ago.
If you are at a school like Harvard your attitude is absurd, as the professors are far above the level of almost all textbook writers, and their lecture notes could be published as is, to form textbooks better than most existing ones.
Years ago George Mackey did essentially this and produced one of the best texts on complex analysis as a van nostrand paperback.
Textbook writing is normally an activity people take up after retiring from research although not always.
Unless you are a very unusually gifted student, it is also likely that your professors really do know a lot more than you do, as most of them are functioning at a level far above the level of the courses they are teaching, and the best ones incorporate this knowledge into the classroom.
Remember, even if your physics class uses the lectures of the great Feynman, he has been dead for some time after all, and thus is hardly current, e.g. on string theory, or quantum cohomology.
Of course your professors may be poor ones, and your class time may be wasted, but then you would be better off getting into a better school perhaps.
There are many people fom whom you can learn profitably, and much faster than by reading. I would say it should take most students several hours to read and absorb what I can tell them in 50 minutes in class. Even longer perhaps to decide which is the most valuable thing to read.
I teach at a traditional state school, but every year that I teach my class, I introduce something that I have not done before, and which is not in our book, although indeed if you read only the best books, such as Courant, you will find almost everything I have ever thought of.
You might try having a conversation with some of your professors. If they impress and interest you, try their classes; if not, try finding some who do.
you are very fotunate to be able to elarn so easily by reading but the best mathematicians learn from other ones mostly, except perhaps Ramanujam. If you are like him, I have nothing to tell you, but then you would probably not be asking either.
