Does computer science need a Feynman or Sagan?

A few postings ago I gave links to data about CSci enrollments.

It's a hot topic: a recent article in Business Week, an interview of six CSci profs in Computer World, and Grady Booch's response on his blog.