I’ve actually seen this sort of thing—claiming a result while backing off from claiming it, circulating a writeup to a few people but not really circulating it, burying an observation where no one will find it—happen over and over again in science, and its invariable effect is to leave fields in a state of utter confusion. There’s an excellent reason why, 350 years ago, science moved from the “announce-by-cryptogram” model to the model of rapid, widespread dissemination of research. And I’m not willing to forsake the attendant gains in human progress, just because some commenters here seem to enjoy the romantic image of someone stuffing the proof of a theorem into a bottle, throwing the bottle into the ocean, then going back to collecting seashells (or whatever), secure in the knowledge that the history of mathematics will need to be rewritten once the bottle washes up on some distant beach a thousand years later. Sorry, not how it works in this civilization.That is yet more interestingness by Scott Aaronson in the comments of his blog. There is a thin line here, no doubt, as too much marketing and not enough meat is anathema to progress. But yes, in order for our results to be useful, we must attempt to ensure that they are heard by people who can use them. One way to do this is to split disparate ideas into separate, shorter papers, even though that practice is sometimes disparaged.
Of the well-subscribed to bloggers I read, Aaronson is one of the few who consistently responds to commenters. He also seems to have grown quite tired of his commenters, as he has threatened to shut down his blog more than once. Perhaps long-lasting success in the blogosphere selects for people who do not respond to comments.