The mean score of Americans with European ancestry is 524, compared to 506 in Europe, when first and second generation immigrants are excluded.
He concludes that the American school system is actually better, if one makes what he admits is a crude adjustment for demographics (with no adjustment for demographics, European scores are better). However, what might happen if you had a sophisticated way to adjust for the ability of children entering school? Would all of the differences between school systems become insignificant?
Thanks to Tyler Cowen for the pointer. As my post on teacher evaluation indicated (note: be sure to read Fabio Rojas' disagreement with me in the comments), I think that there is a good case for extreme skepticism on the ability of teachers, education systems, or schooling methods to make a meaningful long-term difference.