

and that’s just at a glance
That’s a you problem. As it turns out, science isn’t based on glances and vibes.
It’s plainly evident that taking any person with high testosterone and getting them to train at a physical activity will almost always result in better performance than training a person in the same way with lower testosterone.
The papers talk about the evidence based on ordinary, presumably nonspecialised individuals, not cherry-picking a few thousand people who have trained in such a way that testosterone can make its difference over time.
Do you usually interact with people at parties the way you interact with people in the comments section of a link aggregator?