So US baby-names-by-gender is actually one of those areas where digitization has given us much better tools to understand how frequent these sorts of things were. The Baby Name Voyager is a website that lets you track baby names, based on census records, over time, including their gender. So you can put in "Frank" and see that in the 1880s (the earliest it goes up to), the name "Frank" was extremely popular for men, but the number of women named it was slim-but-not-zero. In 1880, "Frank" was the 6th most popular boy's name, but it was the 479th most popular girl's name. Again, that's not zero — but it's pretty uncommon — out of every million girl births there were maybe 150 "Franks."

Now, there are limitations to this, aside from the fact that we are talking about a year 16 years later. We can't say if that's uncommon in a place like Buffalo, and the more general question — would naming a girl after her uncle be uncommon? — isn't revealed in such data. But we can say that "Frank" would be an unusual girl's name, whereas "Frances" would not be (by contrast, it was #42 in 1880, meaning that 3,000 out of every million girls born would have the name; interestingly, a boy named "Frances" is even more rare than a girl named "Frank" — out of every million births in 1880, there would only be 70 boys named "Frances." Change the spelling to "Francis" and the numbers changes dramatically, though).

One of the things I like about the Baby Name Voyager is you can see how certain names have been more gender-neutral over time, and how these trends have changed. So "Leslie" was a boy's name until the 1940s or so, when it very strongly started to become a girl's name. But in this case, "Frank" doesn't seem to have ever been considered something common for a girl to be named.

(And according to the Baby Name Voyager, the number of boys named "Sue" is approximately... zero. However there have been a small number of boys named "Susan," apparently.)