Will have to look up her book some day. This is new to me.
Well that was based solely on observation from people around me. I find that it works well for Is... IJs in particular. I haven't looked at it closely for Es (don't usually interact with extroverts much...)
Using that tends to give you 8 categories, since 2 types share each function.
e.g. ISFJ INFJ or ISTP ISFP
That makes it easier to select their type, since the difference is in their dominant function.
Edit: just re-read what you said and realize I'm repeating what you've said already... (Experiments are getting to me)
The only saving grace with the method is that it's relatively easy in identifying their dominant. Under normal circumstances, the individual will alway use their dominant function is some fashion regardless of what other functions they paired it with. It's impossible to block the dominant. (I tried) So hopefully that'll give you some idea as an outside observer.
How I wish as well there's a method for type like there is for math. It's an "art", I find that I use multiple methods to arrive at my answer.
This shadow approach is handy, but takes a lot of work. Sometimes it takes a while to get the opportunity to see the person's behavior under stress. I'm not the ENTP that'll use deliberate manipulation to experimentally get an answer. Although I suppose if you really wanted to, you can "test" them.