There are two reasons that “People read on the web” is not a myth (by which I mean a widely believed falsehood). Firstly it is not widely believed, and secondly it is not false.
Almost anyone these days, if asked to describe their main behaviour on the web, would identify scanning rather than reading. And they would be right because that is what most websites are designed for. So it is foolish for “usability experts” to express surprise or claim some merit for the discovery that people typically approach a new web page by looking for the navigational elements that they have been trained to expect.
But although it is not widely believed that “people read on the web”, there are in fact many web sites which are designed for deeper reading, and although some may initially aproach them incorrectly there must be plenty who want to read deeply online or otherwise they would not survive. (Examples include thoughtful blogs like PhilosophyTalk, and Real Climate, as well as various on-line books such as David McKay’s Without the Hot Air, along with most of the items linked to by Arts&Letters Daily)
Thus the idea that people read on the web is not a myth, and it is rather the claim that it is one which is the widely believed falsehood. So the idea that it’s a myth is itself a myth which makes it a mythical myth.
(And although it’s not the first on my list I think it deserves the #1 position – which I have been holding open til now while waiting for the perfect candidate).