The difference between books and Internet articles is depth. A good book can take you from knowing nothing to a relatively deep understanding of a topic -- by nature of its format. Consider for example one of the classic books of all time: Newton's Principia. You can start reading that book with just a basic knowledge of geometry, and in a couple hundred pages later you'll know more about physics then you ever thought possible. And that book was presenting physics
for the first time! That's not something you can replicate with Internet articles. It would be much more trouble to do that over the Internet, and you would cry "please, write a book and send it to me so I can read all this information in a easier way". Other books are like that, not only in science but also in literature, philosophy, computing (e.g.: TAOCP) and so many other areas.
The problem we have nowadays is that, because it became so much easier to write books, we have more and more books that don't add much to human knowledge. In the old times, writing and publishing a book was hard, which meant that the threshold for acceptance was much higher. Our society views this downgrading of book quality as a sign that books are not a good format, when in fact this is just an unintended consequence of our technology.