No way, 1-2% is more like the range. For the western world, for serious studies e.g. tracking surnames vs Y chromosomes. Most much higher figures come from counting only disputed cases, which are obviously going to be skewed.
Indeed the Sykes data is long-term history, and I think of a middle-class sample. But it does teach us that received wisdom was badly wrong there, which should make us less confident of the same wisdom elsewhere.
The bone marrow study was 2012, and no class has a monopoly on Leukemia. I would not be surprised if there were a class gradient, but an average of under 2% doesn't leave room for a very big underclass to be scoring 20% or 30%.
These are England and Germany, things could be different elsewhere, I don't know whether there is good data.
