[bpodgursky]

This is trivially disprovable; it's easy to think of longcuts, ways to accomplish or learn something inefficiently.

There would be no difference between a great teacher and a terrible teacher otherwise. Or a great coach and a terrible coach.

[runevault]

The entire software industry is based on creating new efficiencies... On top of all the stuff you said I just... don't see how anyone could believe that.

[m463]

I've always thought research papers were a longcut to comprehension.

I don't know if this is because the subject matter is novel, or that the authors are better at math and science than writing.

I wonder if there could be a standard for grading a research paper on readability to raise the status quo.

[cedilla]

Unfortunately, readability is at odds with other, more important goals of the authors, like shoe-horning in references to all important papers in the field, especially if they're likely to be written by reviewers.

Also, a lot of work must be shown to demonstrate completeness and rigorousness, and that doesn't help comprehension either.

The trick is to read only the segments of the paper that are actually worth reading, which is often only the abstract, conclusion and maybe implementation details.