The SAT is highly g-loaded, and highly correlated with all other measures of intelligence. 450 points on the SAT is a big deal, very significant, especially when dealing with large populations.
For a large population I'd agree that it is significant, because a large population doesn't prep for the SAT. For Ivy class schools I think it's counter intuitively not that big of a deal if one portion of the population hasn't prepped.
Furthermore the data isn't really believable (remember the article used pretty much made up data). Data from actual schools disputes that. For example:
"Harvard's Asian Americans in the Class of 1995 have average SAT scores of 1450, Blacks averaged scores of 1290, whites scored 1400 and Hispanics averaged 1310, the report states. "
"Asian-American students who enrolled at Duke University in Durham, North Carolina in 2001 and 2002 scored 1457 out of 1600 on the math and reading portion of the SAT, compared to 1416 for whites, 1347 for Hispanics and 1275 for blacks, according to a 2011 study co-authored by Duke economist Peter Arcidiacono."
"Asian-Americans admitted to the University of Wisconsin’s flagship Madison campus in 2008 had a median math and reading SAT score of 1370 out of 1600, compared to 1340 for whites, 1250 for Hispanics, and 1190 for blacks, according to a 2011 study by the Center for Equal Opportunity"
And I suspect if you control for recommendations, geographic, and socio-economic diversity the gaps would shrink, not grow.
