To be fair, without knowing anything about the methodology of the study, I assume they're only counting land labor of peasants in those 1620h, and discounting all manual labor for self-consumption such as sewing, basketry, woodworking, and all the infinite tasks of the preindustrial life which were mostly done socially (you'd get together with the gang, weave some baskets and get drunk gossiping having some wine). Then again you could say the same about commutes today, unremunerated house labor, care for infants and elderly, etc.