In praise of repetition?

Just read James Dyson's Wired blog post called "In Praise of Failure", characterizing the trail-and-error nature of innovation.

Overall, I found his characterization of invention dead-on. However, I've noticed that Dyson Labs tends to use the same basic mechanism (centrifugal force) and applies it to all his new creations (e.g. the cyclone and air multiplier technologies). If each of his creations consisted of a novel mechanism, would the failure rate be much higher? Or is it an inventor's process that is key to successful innovation (with a healthy dose of failure, of course)?

It appears that each inventor has their own style: Edison would exhaustively prototype an idea, Dyson uses the "template" approaches, and others use whatever they are most comfortable with. And that might be the most important take-home message of this article: invention is a creative activity, and so it must be done in a way that allows the inventor maximum inspiration.