US demographics surveys typically lump the entirety of Europe (except Spain), most of Russia, and the middle east into "Non-Hispanic White". Iranians would definitely be marking themselves down as such. Occasionally you'll see one that has subdivisions of "White", typically splitting it up into Eastern and Western Europe, and Middle Eastern.

We can debate the various subtleties of "white" till the cows come home, but that is irrelevant to how accurate those graphs are. Let's assume the graphic erroneously counted 100% of Jews as white, instead of 90%. That would change the estimate of non-Jewish whites in the US from 55.4% to 55.64%, and the estimate of non-Jewish whites in the Ivies from 29.2% to 30.63%.

This changes the under-representation of non-Jewish whites from 0.53 to 0.55.

Is this so significantly different that the entire graphic can be dismissed?