If there was an advantage to just being paler – say because of increased vitamin D production – then any mutation that moderately reduced function of a gene in the melanin pathway would be favored, as long as the change didn’t screw up some other function. For that matter, the same is true if sexual selection favored paleness.
You’d see several mutations being selected – different versions in different regions. G6PD deficiency is like this in regions with plenty of malaria – you see a number of different regionally common reduced-function variants, all of which seem to protect against malaria.
MC1R mutations may be like this in northern-western Europe. There are many different function-reducing variants – 9 with frequencies > 1%. Mind you, that doesn’t prove that Vitamin D is the driver, but it is consistent with that possibility.
But SLC24A5 is not like that: there is one sweeping variant, not several regional ones, as you’d expect if the true story is vitamin D. To me, this suggests that a special advantage is conferred by this specific mutation, not just by a reduction in function. That, and the fact that it seems to have been favored by selection in places like the Ethiopian plateau, which have plenty of UV.
It could be that paleness itself really is an advantage in low-UV areas like northern Europe, and that the common SLC24A5 mutation has a double advantage, extra vitamin D production and something else, which might explain why it has been one of the mostly strongly selected alleles in recent human history