Readersarefarmore likelytoacceptyourviewsifyougivethemthechance tothink withyouabouttheevidence.Thealternative—offeringgroundlessassertions—istoexpectthem totakeyourwordforit.
(Page 150) I appreciate this point. I think one thing that can be read from it is that writing can be improved when it is written in conversation with the subject matter or the reader. Few people like to be "talked at," and making claims either with subpar evidence or without evidence at all--while being all around bad analytical practice, as this section explains--can make the writer come across as overly confident and patronizing. It leads to both poorer claims and a worse experience for the reader.
As a silly example, I'm reminded of a critique I saw once of that BBC Sherlock show. Good mysteries give the reader (or viewer) all the clues they would need to logically put together the conclusion themselves. The TV show, however, set all of its mysteries up with data the viewer was never given access to until the big reveal, leading to a pretty terrible watching experience, all in an attempt to make Sherlock appear to be more of a super-genius. I think this can apply here, too.