That would be reasonable start point, wouldn't it? And then if we actually start to observe differences between male and female riders, we would hypothesize that there might be something going on. However, as it is often assumed (also by some members of this board) that there must be a difference and then work their way backwards in order to satisfy their assumption. IIRC there was a paper looking at bet rates showing how folks underrate female jockey performance, which is a bit odd as folks make money with establishing good betting ratios.
My conclusion is fairly simple. There is insufficient evidence that indicates a significant impact of the sex of the rider on the outcome. I have provided at least two references that have looked at it. So far your only counter-argument is that you do not believe it to be the case. Skepticism requires data and so far only one side has provided any.
A no effect finding requires a lower burden of proof. If a homeopathic drug does not shown an effect compared to a placebo, we would not simply assume that the study was flawed and ask whether they used the correct dilution of nothing or whether the flasks had the right silica composition.
If you have any evidence that for some reasons only the male jockeys were sickly, feel free to show it. Otherwise you cannot just selectively dismiss data. Using your approach I could simply dismiss every single study in existence by increasing the burden of proof until I find my bias confirmed. Note that if there is a follow-up that looks at more factors and finds certain associations, that would be a different matter. As it stands, there is no study I could find that contradict these findings.
This does not make sense. Here they decide to go further than other studies (which looked at win ratio and could not find evidence of male dominance) and try to look at it mechanistically. Again, here is evidence and you still have provided nothing. In my world, data is crucial and trumps gut feeling. And as iNow explained yet again, one does not simply start with an assumption and then ignores all evidence to the contrary (or only looks for support).
If I am generous the one supporting factor could be that there are male winners than female winners. But then we have seen that there many more male jockeys than female ones. So that makes a straight comparison a problem. The other issue is of course that horses are very important. If you put a great jockey on a weak horse, it is unlikely to suddenly turn into a winner. So some folks tried to account for horse rankings and then see if women who ride similar ranked horses as their male counterparts perform worse. But the data does not pan out (so we can not simply assume that it is happening). Then perhaps the effects are subtle and maybe men can slightly but significant improve or at least change how horses perform (does not have to be due to male physiology, for example, potentially men might treat horses differently resulting in performance differences). The last study did not find that, either, except for weak significance related to heart rate, which is difficult to translate into performance increases.
So if there are sex specific differences between men and women, they appear to be weak or at least difficult to detect. From this dearth of data to get to the point where we not only assume a difference, but also think that it is due to physiology, is just straight up bad science. It basically has the same level of evidence as claiming that Asians are biologically better suited for maths.