My engineering university offers a relatively new option to do an article-based dissertation, where the primary research is submitted (and hopefully published) in several scientific journals (at my university it's 3). The dissertation is then shorter than a typical PhD, because it describes how the articles fit together to form the thesis, etc.
It's a relatively new idea (for engineering PhDs and for me), which I find interesting as an advisor because it engages PhD students more in the research experience (publishing). Also, it is theoretically more efficient for the advisor and student (as a co-author), since time and energy spent on revising could be more focused on getting publications, and not only on a big PhD dissertation that few people will ever read.
There are other advantages described here (not my university).
My question is not about whether it's good or bad, but how the role of an advisor on co-authored papers might change in such cases.
For example, when students write a traditional dissertation (masters or otherwise), they often struggle with communicating. Students grow and improve written communication and contents of the dissertation in an iterative and incremental process (draft revisions after feedback from the advisor).
In traditional grad-student co-authorship setting, I would take a more active role as an editor (as my advisor did when I was a PhD student) on a paper, mostly because of experience and to increase chances of getting an article published. Sometimes that role is minimal, if only a workshop or conference is targeted, since it might be easier to publish there.
But with an article-based PhD, it seems that the active approach in editing co-authored journal papers is essential, and in effect writing a big part of the dissertation for the student. I realize every case is different.
I'd be happy to know from experienced advisors in this setting to know if and how an advisor's role must change in article-based PhDs.