My objection is that if you run s/random matrix)/magic/g on that article, it is not significantly degraded. I hate most arguments based on "rewriting the original post", so let me be clear that my objection is that the actual information content on the technique itself is not merely low, it is nearly zero. I question the utility of writing an article like this when they clearly don't expect the audience to have even a high-school understanding of matrices. (I mean, the one bone they toss to "what is a matrix?" isn't even correct; matrices aren't necessarily "square"! And again, I reiterate, this is something learned in high school.)