The relevance of SFFA vs. Harvard is to demonstrate that it's possible to have a substantial overrepresentation - over 3x in the case of Asians at Harvard - despite actively discriminating against the overrepresented group. Whites are only ~1.2x more common at Perkins Coie relative to the general population.
You can keep repeating the line that because a company has X% of Y race it must be evidence of discrimination as many times as you want, repetition doesn't make it true.