> Are you sure? Did they really lose 8 hours worth of work? How is "time spent" any better a metric for productivity than counting lines of code?

Coding is a poor example because there is no generally accepted metric for productivity. (Figure out good way of measuring programmer productivity and you'd have a mint.) However, some jobs do have a valid metric. Data entry: 300 forms take an average of X hours to do. After X hours reading Reddit and not entering forms, 0 forms have been entered. Simplistically, X hours have been 'lost'.