It seems they link to a discussion page [1], which references this document (I can't find a published version):

    Computer text line lengths affect reading and learning 
    Peter Orton
    https://cdn.tc-library.org/Edlab/eye-tracking%20article.pdf

But it seems that Peter Orton has published quite a few works in this area [2], [3]. Specifically [3] seems to note that with a less wide text column width readers are not statistically significant faster (but are a bit), but do see that for difficult reading content, the shorter width requires less re-reading then a wider one along with significantly better retention of the contents of the document.

[1] https://www.mediawiki.org/wiki/Reading/Web/Desktop_Improveme...

[2] https://www.scienceopen.com/document_file/671b2d53-73ce-4a91...

[3] https://sci-hub.se/https://link.springer.com/chapter/10.1007...