This is an interesting experiment - seeing whether people respond better to robots who are assigned an ethnicity close to theirs.
It's pretty inconclusive, though.
I'm not surprised people are reluctant to feel empathy with things that look like BINA48, tbh.
BINA48 is sort of a weird corner case, from what I understand. Nobody is really allowed to interact with it outside tightly controlled circumstances, and it's being programmed and controlled outside of conventional research structures.
Sherry Turkle would likely argue that people do better with robots that are on the far side of the Uncanny Valley rather than the near one; MIT's research indicates that the less human you make something, the more we anthropomorphize it.