None of those things are the reason a plane has pitch stability. Most of them combined is the reason. Way back in the past they figured out which were the important details and how they relate to the problem. They worked up a formula that gives excellent CG locations. The formula works for all manner of odd aircraft. Looking at it in it's most detailed version, it looks pretty awesome and scares off a lot of people. But they also discovered long ago a lot of the minute detail really doesn't affect the outcome much, and they boiled down a simple version.

The simple version shows you what the important details really are if you look at it and notice there aren't really many things in it. Good thing about those is they're simple too. You don't have to work out center of lift, or aspect ratios and such. That's really good, because things like the Center of Lift aren't fixed on most airfoils. It moves around in flight. And finding the few important measurements, wherever they are on most planes (even the odd ones), is all you have to do. And it's pretty easy.

If you look at the simplest CG application on the internet you'll notice all you need is wing root chord, wing tip chord, LE sweep, wing span (or half span), the same for the horizontal tail, and the distance from the wing LE to the tail LE. That's just 9 measurements. The wing can be swept forward or back. Same for the tail. The things that matter for pitch stability are few and simple.

I'm sorry but I don't know the "rule of CG as a function of center of lift." I do know that if that's all that mattered, there would be a whole lot more flying wings buzzing about our planet. The size and location of the horizontal tail is really important.

With most models like the OP pictured, and especially with "new designs" that are just altered existing ones, about all the designer needs to do is stay close to the CG that worked before the plane was modified. But if anyone wants to nail down a safe CG range, they really do need to consider the size and shape of the wing along with the size, shape, and location of the tail. And that only takes a yardstick, 10 minutes, and a simple application like are all over the internet.

BTW, the good thing about our models is that even with new designs, they usually aren't "odd". At least they aren't so odd that the simple, old formula doesn't work perfectly. And we've got internet applications that work that formula for us and ask only that we provide 9 simple measurements.