The piece seems to make the most negative assumptions about one workflow and the most optimistic ones about the other. Or rather the assumption that you take the effort to learn the optimal way to use one workflow but take no effort to do the same for the other. Most everything looks great when you take that approach.

Things like making a branch off of another branch being somehow so difficult that no one would ever even consider it. Or that there isn't a Github CLI tool that allows you to avoid having to go to the UI.