I think their results are explainable by the fact that they had everybody reading fiction.

Think about how nonfiction and reference (textbooks, manuals) text is often divided into two or more columns per page.
Ever notice how fiction is NEVER printed like that? Fiction is pretty much invariably printed in one wide-ass column going all the way from margin to margin.

My personal experience bears this out. I have a couple of fiction texts printed in 2 half-width columns (these are bilingual short-story books, with the same story in 2 languages printed in parallel columns), and, tbh, the narrow columns degrade the storytelling aspect. It reads more like FACT FACT FACT FACT FACT FACT, like a newspaper article.

Since smartphones are narrow, they force the same narrow-column format, which will really fuck with your ability to read fiction. 🤷🏽‍♀️ Is my hypothesis.

If they found that the same problems don't happen on tablets, and don't happen with nonfiction on your phone, that would be strong evidence for the lipsypothesis here

A related possibility is that most physical books are professionally typeset with line lengths optimized for comfortable reading. Here's a wikipedia article with examples.

The bilingual storybook is a great example of what happens when the line length is dictated by other factors.

I would also be interested to see if tablet/e-reader users fared better, assuming their device's text/column/spacing settings matched the paper version read by participants.

I have a couple of online ebooks from Amazon and the pages of it get mixed up and skip and are sometimes missing. That would never happen with a physical book. I'm too tired to even deal with it.

+1 to all this.

also i wonder whether anybody had the good sense to flip the damn phone sideways 😂