Over at Zusammen, a post on open notebook science and the least publishable unit drew some interesting comments. Jean-Claude Bradley introduced the term "micropublication", which seems to describe the concept very well.
A follow-up article explores the background and requirements for a workable micropublication system in chemistry.
Many of the points apply to any experimental science. But where chemistry is unique is in the widespread use of chemical structures. Cheminformatics is the central discipline needed to make this happen.
We're starting to see early signs that micropublication could work in chemistry. Consider ChemSpider which is to my knowledge one of the first public-facing chemical database that includes user-created content. While not a micropublication system, it does have some of the key elements. The success of the Wikipedia chemistry project is another indication of real support for the idea of chemistry micropublication. Finally, consider that Chemical Abstracts itself was created, up until the late 1960s, by mainly volunteer effort.
How much of a role will chemistry micropublication play in the future of cheminformatics? Perhaps none. What is clear is that a chemistry micropublication system that actually worked would initiate a major shift in the way chemists create - and consume - chemical information.