Could a serious academic article be published in pieces or installments over time?
Balzac provides some precedent for this in the nineteenth century when he published the first serialized novels [my Balzac site, Robb's biography] . It's easy to see how narrative can be created in installments, but could an academic paper be written in this fashion?
There is another Balzac publishing innovation that is common in blogs and Wikipedia that might help keep blogs from degenerating in quality: frequent revision. Balzac used to scribble changes on the final publishing proofs even after his novels were typeset. As his novels ran through successive editions and printings he would change and improve the novel. Despite the succesive improvements in quality, this must have driven his publishers crazy.
How might various genres of academic publishing be transformed under serialization and frequent revision? If under peer review an author found themselves raising the wrong question, the thesis to be proved as well as the support (reasons and evidence) might change. Collaboration and joint authorship might become more common. Toulmin's model could help to explicitly keep track of changes in the logical argument and support from installment to installment [another Toulmin link, Wikipedia].
Just because blogging software is used to publish academic papers does not mean that the content of the academic article has to generate to the level of the opinionated and highly personal blogs that flood the internet. A distinction also has to made between blogs, which are basically serialized diaries full of personal opinions, and blogging software, which allows serialized web publishing, a potential powerful tool for academic publishing.