Not sure what the "income growth" chart actually means, but if both bottom 90% and top 1% income, although by very different tracks, grew by about 3x in 1917-2012, then something is wrong. Because real GDP per capita grew 7.43x [1]. Logically the percentiles between 1% and 10% from top don't matter all that much, and in any case they must be in between top 1% and bottom 90% -> arrive to about same result although by some middle sort of track.

Any explanation? That chart seems to be a gross misinterpretation of something. Proportion of income in GDP could have only increased since, maybe income of the rich, but still. Where did the 2.5x difference in growth rates come from?

[1] https://www.measuringworth.com/datasets/usgdp