I would say that most research universities aren't in the business of "educating". They are in the business of credentialing more than anything else, as far as

I would say that most research universities aren't in the business of "educating". They are in the business of credentialing more than anything else, as far as the benefit to students. The actual professors are often uninterested (or less interested) in teaching as compared to their own research/career. This isn't limited to just private institutions either. Big public research universities also don't do a great job at "educating". My feeling is that students by and large educate themselves, and learn that skill of learning in the process. Regarding the article's main premise: Harvard's value is its existing network effect, which is a student body that is highly talented and selectively chosen (ignoring the discriminatory biases of "holistic admissions"). That student body makes the degree worth more than degrees from other universities. That's why high performing students continue to go there. If lower-quality students are admitted into the Harvard brand, the value of the degree goes down. That's a cost paid by higher-performing students whose merit gets diluted by lower-performing students. If Harvard just plainly educates more students, it will stop being Harvard since that name's weight and meaning will change. As a fix to all this, what we really need is decentralization of education. That will allow students lower cost paths to education, greater choice in educational culture, and bring a focus back to teaching/learning rather than professors' research areas.