Of course, the writer can try to anticipate questions and capture as many facets as possible, but the more comprehensively one does that---for example by covering both elementary and complex queries from all possible audiences---the more likely that one ends up with a mess of an essay that few would want to plod through.
In addition, the readers, having no chance to pose questions, often ends up having to make a judgement call about the validity of the argument based on the limited case presented in the writing.
Note: a dialog conducted via the written words does not really count, because interaction is still there, albeit less immediate than a direct exchange. And in such cases, the people who read the written record of this dialog would not get so much out of it as the two original participants, again due to the lack of interaction.
If you go to office hours, the professor right there will ask if you understand what they are going on about. They read your body language and see your glazed over expression and go into more detail or try and explain in another way. By the time you leave you are far more likely to know your answer and more.
My job now is a research position that is pretty collaborative, and being able to have face to face discussions with my co workers is a huge boon for my own understanding. I think if I were to work remote or communicate via slack I would loose so much understanding for the sake of working in sweatpants on my futon.
What you are doing alone is maybe "silently pondering" but it is just, at best, a tidying up of what you have gained socially.
The kind of thinking we manage to internalise and perform individually is an impoverished kind, and lacks the ability to do a lot for us.
All psychological growth is therapeutic, requiring a therapist (educator, coach, peer, friend, ...) to serve as a foil to push you past the cycles of your current safe state.
It's not just the student who's engaged in the teaching process, the teacher is too. And with small enough class sizes, or the "gold standard" of one-on-one tutoring (virtually all teaching methods are compared to this, in effectiveness, from what I've read), the teacher not only transmits information, but monitors the reception, sees how students are responding, where they're having difficulty, where they're readily grasping the information, when they're going off on false tangents. And knows how to respond to those circumstances. It's often not appropriate to say "no, that's wrong", but instead to prompt the student through the process of reasoning why a thing is wrong.
Socratic dialogue is one variant of this, and is in fact, almost wholly that last element: letting the student wander off into the weeds and either come to the realisation themselves, or have pointed out (and Socrates employs both methods) that this is pretty clearly ridiculous.
The point of dialogue, and of the clapping metaphor, is that it is extraordinarily difficult to dialogue alone. There are some tricks to this (rubber-duck debugging, say), but having another interested, intelligent, and motivated partner almost always helps.
The direct transmission model of education, just play a tape or teletransmit some remote lecture to satellite classwrooms, fails largely because it lacks interactivity. At best you can hope to target the lecture to some preferred band of the group, but you're going to lose those below that level, and bore those above. It's not a completely wasted effort, but it's not nearly as effective as discussion in smaller groups.
What you'll find with most universities that employ large lecture halls is that there is the lecture, given to 100s, possibly a thousand or more students, and the discussion section, usually lead by a TA, with 20-30 participants. In theory this splits the transmission and interaction, in practice, the sections only very partially make up for the mass medium of the lecture.
My own best learning experiences, regardless of topic, have come from smaller classes, typically of about 15-20 students. Even topics in which I had virtually no prior interest became fascinating with small size, good instruction, and an impassioned professor. (By contrast, several classes I'd really expected would be interesting proved highly disappointing.)
Books carve something of a middle ground, as do recordings where the individual has playback control. You can treat a book as a series of inquries and skip directly to the sections of interest, re-read, speed or slow the pace, annotate, etc. (Much of this is described in Mortimer Adler's How to Read a Book.) With video or audio, increasing (or occasionally, decreasing) playback spead, pausing, and replaying, can give some of this capability.
There's also a point in learning in which there really isn't an established dogma or canon, and where the process of learning is actually a process of discovery (we call that research). Even here, without someone who's specifically expert in the field, a set of fellow-travellers, familiar or not with the route, can help in its exploration.
Two hands work together. One hand merely fans the air.
"Great master" he said "You are wise and know many things. But tell me this: What is the sound of one hand clapping?"
The master slapped him.
At that moment, the student attained enlightenment.