Instead of fixating by what degree it's off the question is around the data's merit and it's real world applicability. If it isn't, then it doesn't matter if it's 100, it's uselessness is absolute. If i know for a fact that the chicks around me will break physically even before threatens to twist their arm then it doesn't matter what some bs study says about pain tolerance. Just as an eg.Data interpretations aren't "off" when divergence is by a few factors ( ie, 2x,3x,4x etc). It's absolute. This is as asinine as saying India has higher rape per capita than the West and Indians just report less, when rape per capita in the west is 20x more than that of India.
As for chanakya, I don't blindly follow anyone. Even chanakya got a few things wrong and I can tell you what they are. Coz I have read him and analysed him.
The reason chanakya favoured women only guard for the royal guard of the most powerful empire of its time, is because he saw the 'Brianne of tarth and renly ' scenario: women of lower station treated exceptionally well by a man of high status tend to fall in love with them and you will always protect the ones u love better than the ones you don't.
He saw the main job of royal guard to not be sword saints but those who will keep an hawkish eye out for poisoning attempts or throw themselves in line of fire to protect their high value target.
So he saw women as better fit for this role and specifically instructs emperors to seek a female royal guard, pick them from the lower classes and shower them with love that they are effectively protecting their husband, not their leige.
Whether you agree or disagree with him, this is the rationale he presented.
He may have been right in this context but he also had his reasons when he said women are inherently deceitful and untrustworthy or whatever else he felt needed to be done for the sake of internal security. Now you can agree with one and disagree with the other or agree with both and say they are true in certain scenarios instead of being universal. We are free to acknowledge it without wholly accepting it.