I’m not sure really. Most DJ mixers nowadays use VCAs in any case so they probably use linear faders for everything as they’re cheaper (and since no audio ever passes through them they probably get away with using the shittest ones they can find).
Thinking about crossfaders, it’s almost the opposite of a pan pot (pan pot takes a single signal and sends to two buses, crossfader takes two separate and different signals and combines them to a single bus). With a linear panning law, you’d get a 3dB dip in the middle (which is why they aren’t used in desks). Not sure whether the same logic could be applied to a crossfader though – the fact that the two signals are different probably means it can’t. Maybe someone with more sound engineering knowledge could put me right?
Actually, looking at the datasheet for a Penny & Giles crossfader, they use a linear law.