It seems like the author picked an arbitrary variable to make a point without enough subject knowledge or citations to back it up. Just visiting Argentina would make it glaringly obvious that the education hypothesis doesn't explain anything.
I think it's unfair to focus on the lack of details in Glaeser's blog post, when it links to his 52-page paper[1] (which the post is a quick summary of). In the paper, he writes: "the Buenos Aires data suggests that less than one-half of the population could both read and write in 1869. By 1895, the next available data point, the literacy rate had shot up to 72 percent, which still meant that a substantial portion of the population was unable to either read or write. It isn’t until 1939 that more than 90 percent of the population in Buenos Aires is literate."
It appears that he is focused more on the historical conditions of Argentina's education than on present-day conditions.
[1]http://www.economics.harvard.edu/faculty/glaeser/files/Glaes...
How many other variables did Glasser analyze? And more importantly, how many variables from 1900 can we even quantify? Education is closely correlated with a number of other variables - quality of governance, IQ, culture, religion, level of development, etc. But there is no good way to quantify those variables for 1900 Argentina. Education may correlate well with growth, simply because its an excellent proxy for governance/culture/level of development.