Doesn't it seem likely that the algorithm for determining whether something will offend someone will contain a reference to objectification or something related to it pretty closely?

Such an algorithm probably wouldn't work for people from past epochs, who had a concept of offense quite similar to ours, but didn't have a concept of objectification. And it wouldn't work very well even today in my home country (Russia). Linking offense to status seems more robust to me.