The debating tactic he seems to be referencing is that you seem to claim that large amounts of information will be added, while his claim is only that some information will be added ("don't think he forswore the Hayekian idea entirely" seems pretty weak). It also feels strongly like you are attacking the Hayekian idea, but not providing an attack; saying "I don't believe this" but not substantiating.

In case people are falling to representation bias from your example, here's a way the prediction market could work out:

One extremely ardent organizer or supporter from the town feels that with the amount of popularity she's experienced the candidate has a much better chance (perhaps due to selection bias from seeing the signs and speaking to reporters) than most predictors, she would see the prediction market as a better investment. And if the prediction markets were national in scope but more localized in specialized information (perhaps consider the recent Massachusetts senate race?), a local might have substantial reason to believe that they're local information would be better than national info.