I can't read your link. It's my impression that two-worker households are much more common now among the poor, in which case the poor are working more on the whole, and they're certainly being paid a smaller share of the value of their labour (see the worker productivity vs. income graph).
There is no relationship between "worker productivity" (real GDP/worker or real GDP/hour of labor) and the value of labor. When capital substitutes for labor, it reduces the denominator in the GDP/worker equation, thereby increasing productivity without increasing the value of labor.
Further, most graphs of the sort you mention compare mean productivity (you can't measure any other kind in aggregate) to median income, which is a statistical fallacy.
If you have data suggesting the poor work more than in the past, please present it.
