In reply, I said that I'm not contradicting but radicalizing my standard advice. Killing off the author has always been a way of making our writing be about the facts (or thoughts) themselves, not about our authority, or our position and status as authors. Remember also that the allegedly post-modern thesis of "the death of the author" is actually a radicalization of the entirely modern practice of "new criticism", focusing on "what is actually there on the page", and eschewing the intentional fallacy. Perhaps it's not so much that my utopia is postmodern as that my postmodernism is utopian.
Earlier this year, I tried to make an argument for a utopia in which most academic communication happened through blogs and wikis, entirely free of charge, and with no publishers, peer review or other editorial oversight. But it's important to point out, as I did already there, that there's still room in this vision for what Jerry Davis calls a "curator", which would be made of up traditional journals, with editors and publishers. Since they would mainly present results that are already known (through the free network), however, these journals would need to guarantee a very high standard of writing and reliability. Also, they would need to take post-publication criticism very seriously. That is, it would really say something about a paper that it was published in, say, the Administrative Science Quarterly, and has not been retracted for ten or twenty years. But if this is to mean anything, then we really have to treat the stuff that is published in the high-end, "premium" journals, not as work to be merely "believed", but as work to be criticized and replicated. That is, getting published there should also come at a higher risk of having to retract it, or at least of having to acknowledge your mistakes publicly.
That is why, in my utopia, "merely" critical papers would be at least as important as the original empirical studies. Something like this is implicitly acknowledged every time a journal corrects or retracts a paper that has been found faulty in some minor or major way. This recently happened at ASQ, in fact, when a "concerned reader" found something amiss with a paper on CEO narcissism. It was worth the time that a reader, an editor and an original author had to spend to find and correct the error to make sure that a falsehood wasn't circulating in the literature. In my opinion, the same sort of reasoning should be applied by in evaluating a "critical essay" that merely corrects errors in an "original" study that has already made a co-called "theoretical contribution". If the paper is flawed enough, its contribution will presumably have been to mislead us. Correcting it is therefore as much of a contribution as the original paper had been believed to make.
In short, I imagine an academy where there are a lot of researchers, but fewer scholars. ("The scholar disappears," said Martin Heidegger.) These scholars would be the proper "authors", i.e, they'd have the authority of knowledge behind them. When they said something, they'd therefore have their "names on the line", just as the journals who published them would have their reputations at stake.