I'd be curious to see this smoothed out over a few years by individual. For example, in silicon valley, it's typical to have an occasional stock windfall year

I'd be curious to see this smoothed out over a few years by individual. For example, in silicon valley, it's typical to have an occasional stock windfall year (due to, say, an IPO). If 10% of people have one of those every 10 years, then that ends up being the top percent income in these analyses. However, it could be that zero percent of people make that in an average year (or that 0.99% make arbitrarily more than that in an average year).