Academic educational programs have exactly been the source-without-a-scalable-distribution-method for at least the past century.

You're unlikely to find a treatment of mathematics or physics or even Computer Science that is as well-curated, free of coercive bias, and well-presented as it often is in the undergraduate programs at universities that care deeply about their educational programs. (Such universities and colleges do exist; they unfortunately also tend to be expensive, selective, and not always well-represented in top N lists -- especially in the US.)

The reason the distribution method of university education is lacking has more to do with economics than anything else. Hiring truly high-quality people to teach small groups of people difficult content in a rigorous way is expensive. If you skimp on any of those features, the quality of the end result goes way down (c.f. typical MOOCs and the university courses to which they are purportedly equivalent, in pretty much any dimension.)

> There were things that claimed to be streams of pure, refined truth. Those claims didn't become false. They always were false.

Again, I fail to see how this critique applies in any meaningful way to high-quality undergraduate education programs in hard sciences. Nothing is perfect -- and in fact I doubt any of those programs ever claimed to be "streams of pure, refined truth". But they come far closer than your comment seems to suggest.

Now, going back to the contents of the article, universites certainly aren't "streams of pure, refined, and new truth". Such streams very likely don't exist.