This approach is correlational and interpretive. The same raw data points can be assembled into substantively different, even diametrically opposed, narratives depending on which mechanisms, confounders, or weighting someone wants to emphasise.

Correlation is not causation. Just because they appear together doesn't mean the cows caused the marriage system. It could be that a third factor, like high male mortality in war prone herding societies, caused both.

She ignores polyandry (one woman, multiple husbands) which occurs in some herding societies like Tibet. If "livestock = polygyny" was a hard scientific law, Tibet shouldn't exist as an exception.