The linked article seems to have a predetermined conclusion to which it is trying to fit the data. Some examples:
Scare quotes are used to implicitly treat "reform" and "reformers" as somehow less than the words imply, without justification for this disdain.
Loaded terms are used to color perceptions of various elements in the plot ("beautiful" kids, "brutal" history, "tragedy", "painful" scores).
Why don’t American kids score at the top on international tests? Our brutal history is part of the answer, as is the immigration policy we maintain so people like Matthews can pay low wages to the people who care for their homes.
No evidence is provided for the claim that history or immigration policy are the cause for the discrepancies in scores.
Basically, the piece seems to invoke "think of the children," quote some opponents with unsubstantiated outrage, and call QED.
The fact that's totally ignored in this article, and which I find most interesting, is that American students of a given race outperform their peers in their native countries. The next question to ask is whether this correlation is meaningful, and if so, how to use it to improve education in the future.
http://www.dailyhowler.com/dh122410.shtml#MINION
If you would like to verify his data, the source is here:
http://nces.ed.gov/pubs2011/2011004.pdf