Speaking only for myself, not co-authors: I think the concept of personhood has become highly politicized in the US, due largely to abortion laws that attempt to limit reproductive rights by conferring legal personhood on fetuses. Medical organizations have come out strongly against this, e.g.:

The American College of Obstetricians and Gynecologists (ACOG) opposes any proposals, laws, or policies that attempt to confer "personhood" to a fertilized egg, embryo, or fetus. These laws and policies are used to limit, restrict, or outright prohibit access to care for women and people seeking reproductive health care, including those who are pregnant, those who are trying to prevent pregnancy, and those who are trying to become pregnant, and they have been used as the basis of surveillance and prosecution of pregnant people.

Our results suggest that US bioethicists overwhelmingly believe that abortion is ethically permissible, and it's thus possible that their responses to the "A being becomes a person..." question were influenced by their views on the permissibility of abortion and wariness about how concepts of personhood are being used to restrict access to reproductive care. 

(Separately, you may also be interested in this recent paper.) 

Feel free to DM me your email; I can send you a PDF!

Speaking only for myself (not coauthors), I agree it's a surprising result!  That said, I think their position may be more nuanced than is evidenced by their responses to this question alone, because their responses to a related question are as follows:

When deciding which patients are eligible for an expensive treatment, it is permissible to consider:

• The patient's expected post-treatment quality and length of life (71%)
• The patient’s expected post-treatment quality but not length of life (8%)
• The patient’s expected post-treatment length but not quality of life (9%)
• Neither the patient's expected post-treatment quality nor length of life (12%)

I think the apparent discrepancy between respondents' answers to these two similar questions may be partly explained by the fact that people found it difficult to choose between, e.g., a 25-, 10-, and 1-year-old, and so basically treated "equally important" as "unsure." To put this slightly differently, I suspect if we had asked "Is preventing a death equally important irrespective of age?" and had given "yes" and "no" as answer choices, a much larger percentage would have said "no."