It does work. The amount of GDP that the US economy gained from 2009-2011 is some positive value. If you look at the share of GDP gains that was captured by the top 1% earning households over that same time period, it's 121% of that value. I can see why you say the semantics are messy, but it is a sound analysis.