Because if you don't define a thing then you can't really make any definitive statements about it.
I've been researching branding books because I'd like to write one on brand. But you know what? Most of them are fluff, and the one's that aren't fluff are deadly boring and researched to death. The only books I really like are the ones written from experience. Because they don't say, "Here is X" rather "Here is my experience doing X." And that can be a really good story.
In this 2002 educational video I'm watching - "In Brands We Trust" - the narrator says that brands are fascinating precisely because we cannot exactly define them. I think that's true. But it shouldn't stop us from trying. Otherwise your statement about what determines brand value is as good as mine, or anyone else's.
At the end of the day a brand can be lots of things, but its tangible aspect is reducible to cash. What are we willing to pay in order to have the branded version? And what are the elements that go into that choice - how much of it is the culture? the story? the features? etc.
When we have the answer to those questions, backed by research not sponsored by a self-interested vendor, we'll be on the way to treating branding like a scientific field. Which is the way it really ought to be in order to be taken seriously.