It could work more like the open source software project model: you would get your respect from having several "accepted pull requests" to the literature. The difference is that your PRs would be better framed in terms of identifying your specific contribution.
As it stands now, the publishing model is the equivalent of writing your own version of the project -- with no easy metadata about your specific innovation -- and asking it to be accepted or rejected as a whole.
With that said, it's understandable why to keep doing the latter version until academia has standardized on the former.
As it stands now, the publishing model is the equivalent of writing your own version of the project -- with no easy metadata about your specific innovation -- and asking it to be accepted or rejected as a whole.
With that said, it's understandable why to keep doing the latter version until academia has standardized on the former.