That seems to be based just on looking at the percentages of the four major ethnic groups (white, hispanic, black, asian). The study doesn't look at other axes of diversity, like nationality, language, etc.

I lived in Atlanta for a long time, which is very "diverse" in the sense that there are large numbers of each of whites, blacks, and hispanics. But it was nothing like New York where you had viable communities of say Trinidadians versus Jamaicans, etc.

Anecdote: On our wedding anniversary a few years ago, my wife and I had dinner at a new French restaurant not far from our house. The server looked vaguely Asian, but she spoke English with what sounded a bit like a French accent. I asked if she was Vietnamese. She said she was from Kazakhstan. Surprised, I said I didn't realize we had Kazakhs in Houston. To my shock, she said, "there are about 5,000 of us here."