There are disadvantages to RPN as well. I read a book on the Forth programming language (also uses RPN) and some calculations that are algebraically simple require pretty bizarre stack shuffling with RPN stacks. I wish I had the source example. It was something pretty standard in a highschool algebra book like the midpoint formula or something like that. The reverse might be true as well in some situations, but I can't recall any examples outside of RPN's historical advantage against old TI calcs in the 80s. Maybe a forth or RPN enthusiast could help.