In all studies, we made certain that the participants and the people in the images were from the same nationality, since cultural familiarity is critical for the face–name matching effect to occur.
Additionally, this survey was conducted by Israelis, and since it says it was translated into English in the paper, I assume it was conducted in Hebrew. They say “socioeconomic cues such as age and ethnicity are experimentally controlled”, but I don’t see that they explain how. My suspicion is that the results are affected by non-facial cues like clothing, hairstyle, facial hair, and indeed age. For example, if I showed you a picture of an old woman and asked if her name was Doris, Helen, Megan, or Kayley, which do you think it is? If I showed you a picture of a guy with short dark hair, possibly graying, beard stubble, and a collared denim shirt, is his name Edgar, Clarence, Emil, or James?
Further, since they did some kind of control over the prompts, I have to assume they presented faces and names the respondents would be familiar with, meaning this does not necessarily hold outside of Israel and Israelis (and I assume mostly people ethnically Israeli Jewish). This reinforces my belief that their methodology is flawed, and while people might look like their names, their faces themselves do not change to fit, rather there’s a correlation with other factors like age (i.e. name popularity over time), grooming style, and so on.