The dataset doesn't ignore PPP, it's expressed in "international dollars" which is a standard PPP adjustment to US levels. That's not really discussed because

The dataset doesn't ignore PPP, it's expressed in "international dollars" which is a standard PPP adjustment to US levels. That's not really discussed because it's an utterly standard thing to do in cross-country comparisons, but the article might have been better if it talked about the fact the data is PPP adjusted so the actual dollar amounts earned in the poorer countries are even lower, and the fact that PPP adjustments aren't perfect so people on $6.70 of imputed US purchasing power can still pay rent