He grossly overstates the stability of subscriptions and their predictability. They _can_ be more stable than advertising revenue when faced with some kinds of disruptive factors... anyway, written from the grass-is-greener-on-the-other-side-of-the-fence perspective of someone who hasn't been there.