I think the interesting finding is that the biases in sex ratio were towards the more successful sex. That is, if A has more sons than daughters, each of those

I think the interesting finding is that the biases in sex ratio were towards the more successful sex. That is, if A has more sons than daughters, each of those sons has more offspring on average than other males. But if A has more daughters than sons, each of those daughters has more offspring than other females. That implies that the bias is adaptive: whatever is causing it, it is maximising the number of grandchildren A has. Also, I don't see any problem with conjecture in scientific papers.