Oh crikey.. there's a question!
Spotting patterns are actually among the most ancient - there are cave drawings of spotted horses. In spite of that, it occurs in relatively few breeds - I suspect (but this is just my theory!) that it's another form of camouflage, and so developed in those breeds whose territory was of a type where that pattern would be useful in hiding them. In places like Iceland adn Shetland, spotted horses would probably stick out like a sore thumb - the land there is much more suited to the dun/dilute type camouflage you see in those breeds.
As for what gene(s) it is that causes spots... that's still largely a mystery. It's not one gene; rather, it's a complex of genes that give all teh different shades and patterns (bit like sabino in that respect).
All horses with spots will have at least one copy of Ap (the Appaloosa spotting gene). Leopard patterned horses are a good example of this - as are those with spotted blankets (Ap is responsible for the spots, not neccessarily the blanket). It also causes mottled skin, striped hooves, and the sclera round teh eye seen in Appaloosas.
There is another gene that has been identified in theory - the snowcap gene (Sc). That casues the blanket - which can be frosting over the hip, a blanket, or cover the horse from wither to tail with a white blanket. A horse must also have at least one copy of the Ap gene for the blanket to appear. Blankets are visible from birth; Ap patterns occasionally are not, so a horse will develop teh mottled skin, sclera, striped hooves and spots later (hence some Appies sold as 'will spot out'). Horses homozygous for Sc are 'fewspots' - mainly white horses with a few very small spots in essence. Fewspots will breed blanket patterns to horses that don't have snowcap, providing they also ahev at least one copy of Ap.
The genes that cause spotting patterns are known as the Appaloosa complex of genes. Snowcap and Ap are incomplete dominant genes, and it's been suggested (by Gower - who had lots of amazing theories but whose work sadly also sometimes lacked evidence) suggested that it's the combination of those two genes, and two others she theorised about, that cause such a huge variety of patterns. One of those is varnish roan - that gene acts in a similar way to grey, gradually replacing the original pattern with a distinctive roan pattern. The other theoretical gene was one responsible for dark spots, the leopard gene - responsible for Leopard pattern Appies.
Whether or not there are four distinct genes, or more, is still a theoretical matter.
It can be predicted to an extent what patterns will throw what, but the actual variety is still a bit of a mystery. Much like sabino in that respect!
