this seems to consider "text being read after formatting" and "text being read before formatting" as different things.
Which I guess, if you're the sole author of the text might be true.
But in my experience most text that gets rendered is also read and edited by multiple people in its source form, so why wouldn't you want to make source just as easy to read?
I think it's about optimizing for different types of reading. When you're reading the final text, you're reading to absorb the content. When you're reading the source text, you're reading to find edits you want to make. Using more line breaks is a way of making the document easier to scan if you're familiar with the "shape" of it.