I think mean-variance can be used to illustrate fundamental principles elegantly. It’s just that it then usually proceeds to studying all its details, pretending that it is sufficient and not progressing to something better. This is the essence of my critique