Well I see their point, but I also see possible flaws. While workplace accidents seem likely to correlate with drunkeness or haste, I can see how they might be different too. They are looking at averages, accidents across the whole plant. Maybe the rate on Monday woudl be low because everyone is rested, but those that do occur are often because of hangovers, and they raise the rate to the Tu-Th rate. On Friday, maybe the rate would be low because everyone is in a good mood, except for those whose good mood turns into haste, and the average of the two together comes out about the same as Tu-Th,
Or maybe accidents and faulty workmanship are not that closely related. Maybe even people with hangovers care about their own safety and can avoid accidents as well as those without hangovers, but they don't care about quality. One is about physical coordination and the other is about attitude. (I don't know, I've only been drunk 3 times and I had litle or no hangover each time.)
Fallacy towardds the end: " may just turn out that most workers don't get totally wasted on the week-end". I don't think anyone thinks most workers get drunk on weekend or careless on Fridays, only enough that if there are assembly defects, they're likely caused by hangover or haste.
From what I read that was/is definitely true in the Soviet Union/Russia, in auto-making and other stuff.