Protopersonas are another poor idea and worthy of a separate article.

Why would we guess or assume who people are and THEN interview people? When you start with your assumptions, too many people (especially non-Researchers) do the interviews that validate what we assumed. Tail wags dog.

If you are going to do good qual research, don't do proto anything first. Be agnostic. Don't go in with assumptions about who people are.

I recommend against traditional personas, proto or otherwise. We should be focusing on behavioral typologies or Larry Marine's user profiles.

Cowan's work might be better than nothing, but consider that there are even better ways when we care about quality and outcomes.

And this article isn't about him. It's about these types of methods under various names.