A couple of things struck me right away.
Though the articles in the section do discuss recent research showing the processing demands of reading complicated texts, they do not reference the equally important research (reported last year) showing how a shift from reading longer texts to accessing information and learning mostly through other visual modalities changes the physical brain. (In other words, if you learn to do a procedure by reading about how to do it, your brain adapts to learning that way--and to the level of reading difficulty--but if you learn to do a procedure by manipulating a joystick on a computer and watching the result, your brain adapts to that.) In previous articles on education in both Science and Nature, the changing classroom environment, and the expectation of students that material would be presented in non-text fashion, were discussed, along with the implications for literacy--the ability to pull meaning from text.
Catherine Snow's "Academic Language and the Challenge of Reading for Learning about Science" offers two examples of writing about torque, one from an online forum for low-rider enthusiasts and one from a textbook, contrasting "more informal" with "more academic" language. Predictably (in a journal such as Science) Snow treats academic language as necessary for communicating complex ideas with precision. The authors of other articles in the section also take academic language as a given and students' learning to read academic language as a primary goal.
But as a professional writer and former science/math tutor, I see it differently. Much so-called "academic language" is simply bad writing. If the purpose of writing is communication--and the purpose of academic writing is communication of ideas and data important to "academics" (in this case science academics) then you'd expect "academic" writing to be easier to read. That it is not suggests a hidden purpose or actual inability on the part of its writers. I think both are involved: the hidden purpose is to distinguish among students--to form a gate through which only those "qualified" may pass. Academic disciplines, like any human group, have always had their boundaries marked and guarded. In addition, academic writers find it easier to follow in the patterns they were given when they were students, patterns that absolve the academic writer from the need to make reading a text easier.
Specifically, the example of a textbook explanation of "torque" is written badly--bad in terms of organization and phrasing--and that bad writing makes it harder to read. Over the years, I've looked at the progression of size and complexity in modern textbooks (teaching geometry, for instance, from textbooks quite different from those we used) and have found that as the use of illustrations and color and "relevance to everyday life" increases (along with the size, weight, and cost of the textbooks) the quality of writing in the actual text decreases.
I read a lot of science. We subscribe to Science, Nature, and two medical journals in this household--have for decades--and I also read science journalism and "popularizations" in other venues. In the past, when I was at one or another university, I had access to many other journals and read them as well. So I've been exposed to both good science writing and bad science writing, over many years, in multiple fields, in multiple venues. Academic writing holds no terrors for me; I can plow through pages of turgid stuff (muttering to myself that this guy or gal really needs a class in writing nonfiction and the editor of the journal needs to use a red pencil far more freely) and understand what the writer is talking about. (I can also see where turgid writing is expressing turgid thinking, a skill that was very useful in grad school journal groups.) In my opinion, the best thinkers are often the best writers, willing to be clear, willing to eschew standard academic-writing style in order to maximize communication with their readers. This not not "talking down" to their readers--this is a recognition that if the ideas and the supporting data and logic are complex enough, rapid and accurate communication requires clarity over formality.
We do need better training in literacy through the years of education--including not only reading text, but "reading" illustrations and grasping the fundamentals of logic--but we also need to jettison the notion that writing in science requires a special "scientific language" and the related notion that reading science is somehow different from reading history or sociology or a well-constructed complex work of fiction. Bad writing in science is not an elevated form of writing--unneeded nominalizations, convoluted syntax, impersonal authority, and other characteristics are simply bad writing. Bad writing makes text harder to understand. (Or, converted to academese: "Those texts generally found to be harder to conceptualize are characterized by substandard use of syntax, passive voice constructions, overuse of nominalizations...") If the purpose of science writing is communication...then bad writing frustrates that purpose.