Good point, though that's really just the summary from the magazine. The actual Hart-Risley study (which was mostly the basis for the passage I quoted) shows that the relationship between SES and achievement is fairly linear, as are the relationships between SES and the factors that lead to achievement. I don't have time to scan in the entire appendix of the study, but here is a quick snap from Hart-Risley:

http://www.alexkrupp.com/picture_library/hart_risley.jpg