[Someone]

Knuth agrees with that. That's why he doesn't guarantee rapid replies to almost anything (http://www-cs-faculty.stanford.edu/~knuth/email.html)

And, relevant to the topic: he is quite productive, but I don't think he would make it in academia at his current output rate. It does help that he splits his work into separate papers, though, and that he publishes 'dull' papers about his tooling, too (IIRC, the original plan was to have two volumes of TAOCP. Now, we have 4 plus fascicles plus some books about the software he wrote to lay out his books and create the fonts therein)