Gingras, Yves, et al. "The effects of aging on researchers' publication and citation patterns." PloS one 3.12 (2008): e4048. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2603321/pdf/pone.0004048.pdf
... gives an interesting inside report about the publication situation of academic papers. According to the figure on page 3, the average author of an scholarly paper is between 40-60 years, is called a professor, and has an output of 2 papers per year. The literature which is cited in a paper is between 6 and 8 years old. But can these statistics be true, is the average output really 2 papers per year per professor?
According to another statistics in the academic sector are working 10 million people worldwide who are producing 2.5 million paper per year. With the above cited productivity, around 1.25 million authors are enough to produce such amount of output. And the other are doing assertive work like preparation and formatting of the MS-Word documents.
Let us criticize the numbers a bit. At first I want to focus on the age. It is possible to let people write a paper who are younger then 40 years and older then 60 years? Probably not, it is the same problem like in Wikipedia, that a certain amount of knowledge is needed to deliver high-quality content. I'm in doubt if a student in the first semester has enough wisdom accumulated to write a short paper about robotics. So the age distribution between 40-60 makes sense. A second parameter makes no sense: the productivity of 2 papers per year. This low amount makes only sense, if we are assume that the writer has no computer for textwriting and no internet for doing literature research.
For example, if the only tool a professor has is a 40 years old typewriter and if he writes the first version of his manuscript by hand, then the workflow become complicated. In that case, 2 month for the draft, 2 month for the typewriter rewrite and 2 month for the reformatting at the publisher will be necessary. The amount of 2 papers per year is the consequence.
But, if the professor is familiar with modern technology like the Lyx software, and writes his first draft on the PC without printing it out and without complicated peer-review process he can produce much more papers per year. I would guess that 10 papers per year are possible as a minimum and 20 papers per year if he is working hard.
To make the point a bit clearer, if it is possible to convince the academic professor to change their old handwriting workflow into a modern PC-based workflow, the same amount of people can increase the yearly output to a value of around 10 million papers. If it is possible to extend the number of people who are familiar with academic writing the number of yearly generated papers can be increased further.
Today, the academic system contains of two bottleneck: oldschool workflow in producing content (handwritten and typewriters instead of PC) and secondly a small amount of people who have ever written a so called paper. Both bottlenecks can be overcome with the new Open Access movement and a younger generation which is familiar with modern word-processing software. I would guess the transition process takes a bit, at least 10 years, perhaps more. The general problem in writing a scientific paper is, that it takes many years in learning it. It is no coincidence that the minimum age until a professor gets published is around 40 years. If he has started his academic career in the age of 20, it needs 20 years until he has read major publications and is able to write a qualitative survey over a topic.
Let us investigate possible improvement in future academic publishing. I have identified to possible parameters which can be adjusted: at first the retrieval of documents with Google Scholar and secondly the content production with Lyx. In the past, most researchers have not worked with these technologies. Instead they are visiting a classical library and ask at the desk for a printed journal, and they are writing their manuscript with a typewriter. The reason why they are doing so, is because they have learned so, and are not allowed to unlearn a technophobic behavior. They are not rejecting Google Scholar because of it's weak quality, they are preferring classical libraries because their peer-group is doing the same. To make the point a bit clearer: if a professor is using Google Scholar for get an overview over his topic, and if he is using Lyx for writing a manuscript he will be no longer part of his community. It is some kind of academic failure in doing so.
Explaining this surprising social behavior is easy. We have to only take a look back into the age of the first gutenberg printing press and why it wasn't used by the majority to understand the problem. Technically the printing press was invented in the 15th century, but it takes 400 years until the technology was used by everyone. Technically the first TeX typesetting software was invented in 1979 but even today nobody is using it. A delay of 50 years and more until a useful technology is accepted by the mainstream is not an exception it is the normal case.