Thompy
That's an interesting question;
I suppose that most people don't realize that a discriminator control (or a ground balance control) are voltage limiters. When we look at iron, it is one of the poorest (metalic) conductors of electricity that we can find. That's why we can cancel the small iron before other metals. Then next would come some metals that come into the aluminum range. After that, we would find that lead would conduct electricity a bit better than aluminum. Then of course brass and bronze would fit right in with lead. Gold however, has a much wider capacity to conduct electricity. So does silver and copper.
So the end product can be different with an XLT compared to say, an M-6. There is no way to match the two, save for their behavior in the real world vs the theoretical world.
The answer then, is that iron alone is not the only thing that determines where and how a detector should be ground balanced. Even sodium and magnesium and high copper can affect the ground balance, depending on how much of it there is and whether it is widespread, or here and there, or in little spots in this place and that place. The faster the rate of retune to a predetermined ground balance set or preset parameter, the better the ability to continue cancelling whatever high mineral there is in the ground at each given. nanosecond and at each spot.
In short, ANY metal or mineral can affect the ground balance capability of a detector, but iron minerals are the most predominant found in the ground, and that is what the detectors have their platforms built on. Additionally, iron is the least conductive of all metalsand is therefore tended to as a starting base for discrimination and ground balancing (both are the same thing).
Some detectors have had their scales calibrated to enable certain electroconductive metals to fit where they don't belong on a gradient scale. We see this when some detectors accept aluminum clear up into the high coins range (such as many Minelabs), while some detectors shut it off way low and just above the iron range. This is why many MInelabs cannot completely cancel all aluminum pulltabs.
The problem is, that with too high of gain (such as many Minelabs, some 1200 series Fishers, some high end Tesoros, Many Garretts ie; 100 to 2500 series, etc) the circuitry is overdriven (with ridiculously high gain, actually warping the RMS voltage constant) far beyond the abilities of the detector to combat that - in addition to being able to cancel the high iron content simultaneously.
White's and others have some similar problems but their problems are not as acute as with other brands and some of their detectors. And beside that, there is no constant either, meaning that every White's, Minelab, etc detector are NOT just like their cousins; ie; x-terra vs Expl SE.
So yes, it is (mostly) the magnetite or hematite or pure Fe (iron ore) that affects the ground balance, but that's not the only thing that does.