Tetrapods are rare. Ballpark guess: about 2% of animals (by taxa) are tetrapods. The number is much lower if we go by number of individuals, much higher if we go by biomass.
We might instead ask: Why are all of the large land animals tetrapods? It’s probably just dumb luck. Above a certain size, being a land invertebrate just doesn’t work, at least under current conditions on Earth. So, to be large you have to be a vertebrate. Vertebrates happen to have evolved four limbs early on. With the exception of myriapods, the number of limbs is a “hard” thing to change. Myriapods have large numbers of legs that are almost all the same, on a large number of body segments that are almost all the same—changing the number of legs, or the number of body segments, does not make much difference to a myriapod.
Apart from the myriapods, the greater diversity of limb number in invertebrates is an illusion—we’re not comparing equivalent units. We decided to dump most of animal diversity in one bin, and a little bit of animal diversity in the other bin. The bigger bin has more diversity in it—in limb number as well as in everything else. Suppose we measured the rate at which limb number changes per reproductively isolated lineage per unit time. First, the number is for all practical purposes 0. Second, the number would not be noticeably higher in invertebrates than vertebrates. Third, almost all of the changes in limb number would be losses.
The rate at which limb number changes in lineages is basically 0 because the genes that control limb number and differentiation between limbs also control more or less everything else about the organism’s body plan. You can’t just tweak the number of limbs without also changing lots of other things. Somewhere in that grab-bag of other changes is almost always going to be something that’s simply lethal. The changes in limb number are almost always losses because that’s the easiest pathway to change limb number without lethal consequences. Reducing the number of limbs might happen by basically turning off a set of genes in one part of the body, which means those genes can still keep on doing whatever they’re doing in the rest of the body. If you’re increasing the number of limbs, that’s more likely to need some genetic changes that will interfere with other functions of those genes—the genes will need to be “doing something different”, to some extent, rather than just being turned off in one context.