This is just an armchair statement. I’ve taken a number of martial arts, but most of them only at an introduction level (for a few different reasons), so I’m not an expert.
Any martial art which has real potential to maim or murder your opponent is, likely, going to operate on a somewhat similar system as Rock, Paper, Scissors. That is to say, if you’re given two people of similar physical capability and martial knowledge, then the odds of one winning over the other should be even. One of them will win half of the time and the other will win the other half. But the reality is that one of them will be “better” and will win more often. The difference between them is that one of them is better at predicting what the other will do, because humans aren’t random.
We see the same thing with race car drivers. By all logic, it should just be a matter of driving the perfect line consistently. And if you get out a bunch of pro drivers, and put them on a track all by themselves, they’ll all do just that and clock in effectively the same time. Yet, if you race them all together, one guy will win far more often than all of the others because he understands what they are thinking, he reacts to the other cars better and he is able to set up situations that throw them off of their game.
So while two computer components will always come out even in a game like Rock, Paper, Scissors, between two humans you’ll usually get the same winner out of a series of tournaments. It’s just a matter of knowing the limited options available to your opponent and - regardless that the options are all in theory equal - being able to predict which they will go for, due to human fallibility. In a sword fight, it might be, “Is he going to go for my face or my arm?” In a karate fight, it might be, “Is he going to kick or punch?” The person who can predict this more often will win, even though otherwise they might be completely even in terms of knowledge, physique, and technique.
Different martial arts were started by different men who were better at winning than others. When asked what they were doing differently, they had to have an answer. The guy who could clearly articulate his thought process and teach it to others is the guy would earn the most students and the most rewards from the local lord or king. A guy like this is probably also the one most able to think of better ways of practicing and, based on whatever it was that he said to describe his thought process, is going to try and come up with new moves that follow that philosophy. Stasis will allow the other schools to figure out what the new school is doing and figure out how to counter it. But over time, as the philosophy becomes a central aspect of the school, the techniques will drift away from what they were doing before, and end up being its own art. How well you compete against those other martial arts will cease to be relevant.
But, at the same time, the fact that each art has its own philosophy behind it will make it more suitable to different situations and more appealing and more natural to certain individuals. If the philosophy ‘works’ for you, it might be that your mind works the way that the founder’s did. His movements and the nuances of the thought process might make more sense to you than it does to others.
There’s a former air force pilot who invented a military philosophy called the OODA loop, which basically says, “Between attacks, step back and think for a moment, before moving back in.” While perhaps not that mystical, this statement still allows for better guidance on how to get into your opponents head than if there was no such statement. It gives teachers a criteria on which to judge their pupils and try to get them closer to a learned duration for thinking. It gives them a vocabulary for discussing the thought process of the battle, which they can use to review and collaborate with their students. Minus that, and you’re just at, “Uh, kick his butt and don’t get yours kicked.”
Granted, a lot of it is probably mystical woo, but there probably is some core bit in there which is actually useful for the purpose. The OODA loop would be an example that lacks the mysticality.