I recently gathered data for what I think is the best measurement of this: hours worked per year per working-age (15-64) person. It has risen modestly, from 1,046 in 1964 to 1,156 in 2013. Maxima were:
OECD Data puts the number at 1,790 currently vs. an OECD average of 1,765.[0] Differing underlying populations (i.e. the way you cleaned up your dataset per your definition) probably explain the discrepancy.
Out of friendly curiosity, what're your data sources and methods? The OECD data seems suspect, especially since it has Japan at 1,745 hours/year, less than the US. [Ibid.] That just seems unlikely.
I used BLS data for hours worked and OECD data (supplied via FRED) for population. Presumably your link is using a narrower demographic, since the numbers are so much higher.
Edit: Yeah, that looks like hours worked per person employed
