Is that a reason for not testing though? VO/Chrome is a statistically relevant proportion of screen reader users (about 2% to Safari's 7% according to WebAIM).
What I would say is that if something works fine in VO/Safari and in Windows screen readers too, but VO/Chrome is the outlier that doesn't work ... then sometimes it's fine to say, fuck it that's Chrome's fault, depending on how critical it is.