I'm not really sure what Borgstrom is trying to do here, actually. She's obviously borrowing from computational and so forth models, and it sounds like also from some kind of structural or Information Theory models. But it all seems like a prolegomenon, a setup rather than a result. That is, my reaction is something like this: "Okay, so now I see sort of what your model looks like, so... so what?"
That sounds really harsh, and I don't intend that. I mean the "so what question" some teachers inflict on students (I don't, as it happens, but I'm sympathetic). You get a paper in which the student says, "I have all this great data, and it fits really well into this model, isn't that great?" Yes, it is, but what you've just done is to set up a paper yet to be written.
Let me put it like this. Borgstrom seems to me to have taken a limited data-set for RPGs (I do think, incidentally, that she ought to be clearer about how and why she limited the data in this way, but that's a passing point), then constructed (on bases a bit unclear to me, but that seems to be more a question of what I have and have not read) a model to fit the data into. She has aligned an abstract, theoretical model with a set of (partly constructed) data.
So what's the payoff? She makes these remarks about epistemology at the end, which seem to have bothered some readers; they bother me too, because I don't know what she means by "epistemology" nor in what sense this affects it.
The whole article is so short, at least by comparison to stuff in my corner of academia, that I think she ought to expand by about double. I want to see her apply this model to actual data, not a string of hypotheticals used illustratively, and I want to find out what effect this has for the interpretation of that data. And at present, anyway, I cannot even guess what that rest of the article would look like.
I suppose I'm also tired of projects like this, constructing new models ex nihilo. That's what frp.advocacy did, and Ron Edwards, and most of the Forge material, and various others. At some point we need to see them do something. Classification is a necessary preliminary to scientific understanding, perhaps, but it is not in itself much of an achievement. That was one of my objections to the Big Model, in fact: it seemed to me that you had a lot of fascination with classification and not a lot with analytical implications. Of course, that's in part because what I mean by "analytical implications" has nothing to do with game design per se, thus the fundamental divide between me and Ron.
Borgstrom does something admirable in drawing on academic models to build her structure, and I salute that. But models are built to do things, and this kind of model is (in my experience, anyway) usually intended to do analytic work. So I'd like to know what sort of results and effects this model has for analysis.
To put it simply: what questions does this model help us to ask and answer about RPGs?