Yea this is interesting. I don't have enough evidence/experience to more convincingly argue against this (i.e. the argument being we should teach tidyverse as a first approach). I would love to hear Hadley and co. as well as some of my Stats professors views on this in a head-to-head debate