Damn good point. Now that I think of it, all the classic examples of "anonymous" publication were really pseudonymous. (Publius, et al)
They have different requirements. Votes and cash transactions and similar things require no history, no reputation. They're one-shot actions that should not be linkable to other actions. Pseudonyms are used everywhere in practice, because even my name is effectively a pseudonym unless you have some reason to try to link it to a meatspace human. This is why it's worth reading a book by Mark Twain, even though that wasn't his real name. And it would be worth reading those books even if we had no idea who had really written them. The reuptation and history of the author lets you decide whether you want to read the next of his books. The same is true of academic papers--you don't need to have met me or even to be able to find me, in order to read my papers and develop an opinion (hopefully a good one) about the quality of my work. And that determines whether you think the next paper is worth reading. --John