Two comments about the labor data and political noise surrounding revisions:

1] Establishment survey nonfarm payroll revisions tend to move with economic momentum (so much that rolling revisions look like an economic indicator)

2] The unemployment rate is much less revised.