> the r>g observation (due to Piketty, Capital in the 21st Century) is that both historically and recently the returns to capital exceed the growth rate of the economy.
They do if you cherry-pick your economies, starting points, definitions of "capital" and "growth", and cutoff points, yes.
You have to cherry-pick a lot harder to make r>g go away.
Plotting it vs time makes this clear because you can clearly see what caused what.