First, I'm pretty sure I've seen this article before. How does a 1996 article suddenly become news?
There's a hole in the guy's argument that I'dn't seen before though:
Perfectly operating markets rely on perfect information flowing to all the participants. Approximately perfect markets rely on participants having approximately perfect information. If A is true, it's okay that some people believe A and some believe B; the people who believe A will be rewarded by the market, thus encouraging people to be right. However, if A is true and there is a persistent myth that almost everyone believes that B is true, its a real market killer. In that case, the market will make decisions based on B and will radically misprice anything affected by A or B. (As an example, circa 2006, A = house prices are going to fall, B = seriously falling house prices are negligably improbable.)
Thus if it's true that Dvorak keyboards aren't actually better while the persistent myth says they are, that would be a mark against the perfectly optimal free market. To recover, they would have to argue that, while the Dvorak myth persists, no similar myths exist for economically important realities. They don't do that.
Arguments of perfection really do free market theory a grave disservice. Outside of God's Heaven (which I don't think literally exists) nothing is perfect. Being unaware of your imperfections is an invitation for them to overwhelm you.