One thing that I think these types of channels and content don't do well is show all of the pitfalls and failures along the way. The missteps are
vitally important to learning. This is also a potentially harmful side effect of newly remote teachers moving to pre-recorded lectures.
When you watch a 15-30 minute video(or even an hour long video) and it's highly polished. You do learn how to go through the exact same problem they're working on. What you lose is the methodology of how they retraced their steps when they got stuck. What did they read and reason through to be able to provide their carefully worded tutorial? How did they reason through the misinformation they had in their mind or that they read online?
When starting my undergraduate education I came into the program with a woefully inadequate grounding in mathematics. The calculus professor I had did at least two things right with his teaching that took me from failing algebra courses to completing really in depth 30-40 page projects several times per semester.
1. He had high expectations of those learning from him.
2. He showed his failures by working out problems he didn't work out before the class started. And critically he attached real emotion to those failures.
Focusing on #2:
There was nothing quite like seeing this professor get on the white board and work through a problem from first principles, explain his reasoning at each step... and then SHRIEK.
He would let out a high pitched scream that I can still hear today when he'd notice an error. Then he would carefully show us him backtracking through his entire process until he found the source of error. When he'd finally find it he would circle it and let out an "aha". Then he would continue on back through the problem fixing the errors along the way.
Seeing this process did several things for me as a student: It showed me that I am going to run into problems even if I have a PhD. Basic problems even. But more importantly it showcased the thought processes you need to be able to solve not just the problems you're given, but ALSO the problems you create yourself that come along with a human's fallible reasoning.
So he'd give us these insane Calculus projects to work through and inevitably I'd make some mistake that would ripple through the entire project. But knowing at least he would find the source of the error, check for errors, and being given some of his intuition first hand on where the errors were sourced was infinitely valuable. Without that knowledge I surely would've gotten frustrated and just given up.
Humans need both positive and negative cases in my opinion to learn effectively. Right now we're optimizing for positive cases only. Those are important, but the negative cases are also extremely important.