I've tutored math in Korea, observed the schools in Japan and Singapore, was observing math class in Shanghai most recently, and have used curricular materials from Singapore, Japan, and China to homeschool my own kids in math.
What I observe in the Asian programs I've seen that differs starkly from the vastly inferior curriculum here in my own country (US) is meticulous care put in to the development of:
1) a strategically-ordered, cohesive sequence of topics,
2) explicit list of the ideas that need to be mastered in each topic before moving on, and
3) rigorously field-tested teaching techniques, examples, assignments, and assessments for each idea.
The whole sequence is constantly being experimented with, tested, and optimized like a Japanese industrial, or Apple consumer, product.
In contrast to this meticulously crafted approach to incremental mastery, our US approach is one of each teacher throwing her own bowl of math spaghetti at kids and whatever sticks sticks. Whatever doesn't stick today, well, don't worry, we'll be doing something different tomorrow. Maybe you'll do better with that. And someone else will throw the whole bowl at you again next year, and the year after that. If you run out of years and a lot of things still haven't stuck, well, we did all we could. You're just not a "mathy type" (claim the teachers who couldn't solve a Singaporean 5th grade word problem if their lives depended on it.)
What Sal has produced in version 1 reminds me more of the American way than the Asian way, but that's no surprise. It started off as an act of personal generosity, not a NASA space project. If it stays the way it is, I'll be grateful for the moral equivalent of free pizza. I'll gladly eat some, and I won't fuss that my free pizza doesn't have my favorite toppings.
But if he's planning to go forward, I think the next step might be to back off a bit on the quantity of videos and consider upgrading some of the old ones to a more deliberately designed, cohesive video sequence.
You can combine a well-designed, coherent curriculum for skill training without imposing the crushing baggage of "your worth as a human being depends on the brand prestige of the university you get into."
If you want to train people to be inventive in finding new ways to apply mathematics, and new mathematics to solve problems, then you need a process which leaves some room for questions, creativity, for open-ended challenges, and for some context around mathematics as a creative process.
The industrial production line approach may be efficient, but perhaps requires a lot of external pressure and discipline on kids to keep it on track, and can result in people who are strongly technically skilled within the relatively narrow boundaries of material they've been drilled on, but utterly lacking in imagination or passion for the subject.
In practise I suppose you need something in the middle. It takes disciplined study to develop the fluency to be creative in maths at a given level, but the discipline is a means to an end not an end in itself, and if it's so strong that people are discouraged by shame from asking the questions they need to ask to develop a deeper than rote understanding of the material, then it's got to be counter-productive.
I say this as someone who was almost put off maths for life by the latter approach, but is now doing a second masters in a mathematical field...
I personally think that Khan's response was one of tact and class, as is your comment. Keep up the good work.