I have noticed that you are able to have a "master" qmd/rmd and a "child" qmd/rmd but can the children have children? This does not seem to work.
Additionally, when knitting a document if the "child" is large it is difficult to gage where the process is in terms of creating the intermediate "md" file which is used for the ultimate rendering - you might be looking at "child" computing for most of the time in the terminal. Let us suppose there is a subgroup that fails for the analysis - but - then we wont immediately know for which of the subgroups the analysis failed to compute. Is there a way to get better tracking and progress on this?