I agree that the article plays a little loose with terminology; however, I wouldn't go as far to say that it's a circular argument.
To snitch an example from mathematics, consider formal logic and set theory. These are oft considered the epitome of rigor, enough so that they form "the foundation of modern mathematics." However, when first begining to study these fields, one encounters a sort of philosophical conundrum, "How do you even state the rules if you start from literally nothing?" You can write them on a piece of paper, but without some system of processing, all those rules end up as just ink on paper. The standard terminology for just such a system is "metalogic."
Anyway, at first blush it seems like any such metalogic is inaccesible to mathematical inquiry, but we can use a trick to "lift" the metalogic and logic one layer up. For example, we can (using some metalogic) start with standard Zermelo-Frenkel set theory (ZF), and then ask ourselves, "Is this ZF powerful enough to iplement a version of ZF within itself?" In other words, if ZF is too weak to implement ZF, then clearly out metalogic must be something stronger and more complicated.
Fortunately, it turns out that ZF can implement itself just fine, and in fact, there are much simpler (read weaker) logics also capable of implementing ZF. Said another way, the bare mininum needed to write a program capable of verifying ZF proofs is quite bare and minimal, indeed. Counter-intuitively, perhaps, this line of inquiry has ended up discovering pretty nifty proofs of previously intractible problems. It also has practical implications for proof verifiers and the like (see Metamath[0]).
So, by analogy, I read this article as saying something similar about the metaphysics of our physics. I.e. it turns out that there are some really simple rules capable of generating the complex physics that is the Standard Model. How much can we whittle down the metaphysics? What does that say about our universe?
[0]:http://us.metamath.org/mm.html