But there is a thought process to working things through. Maybe valid, maybe not. It's the one I use.
Bad guy first. How and why would someone threaten you with a gun? Robbery? Intimidate you to a secondary crime scene for something that requires time and privacy? Each of those has very different needs, different geometries and intents.
Under what circumstances would it serves the threat's interest to threaten but not just kill? Does the fact that there isn't just a loud bang tell you something? Is that information you can use?
When and where would this happen? You can't predict perfectly, but sometimes I get the feeling that one of the most common types of armed robberies happen in convenience stores, across a counter. How many practice disarms across a counter? Is there a good technique for that? I can't think of one I'd trust.
There are distances that make a disarm completely impractical, and anyone who has been taught to hold a weapon at retention at certain ranges will be damnably difficult to disarm.
Which all adds up to, "How fricking narrow is the window of opportunity for this technique?"
Then re-examine the question. Because disarming is not the goal. Not getting shot is the goal. Whether that means getting to cover, disarming, creating distance or shutting down the brainstem (none of which are completely reliable) isn't dependent on what you are good at. They prioritize on which will be fastest.
Then, and only then, if disarms are the option, it has to be based on the geometry. That's just biomechanics and any martial artist who has played with another body should be able to see what has the best chance to work. Which motion works with the threat's joints (e.g. it is almost always easier to shove the weapon towards the threat than to pull it away from you). Which motion will get the barrel off your body fastest (I've seen and despised techniques where you pan your own face). That's just a matter of seeing.
Doing drills, almost anything works, provided you do it with full commitment and no telegraph. Action beats reaction very, very consistently. But if you haven't practiced untelegraphed explosive movement... not good.
And this is the part where I rant about technique. I don't like technique dependency. Whether it was the complicated, multi-step locks and handcuffings we were taught at the academy or simple disarms.... grrr.
Here's the thing. I can reliably make the very first action just like I want. Whether that is a drop step pass-parry or slapping a wrist doesn't matter. But people don't react or flinch or anything the same. Anyone who says, "If you do 'X' the threat will do 'Y'" hasn't fought people on meth. Calling a four-step move one action doesn't magically turn it into one. The first action, if you can do it explosively, will work. Everything after that depends on your adaptability. Which depends on your ability to apply principles.