My point is that those metrics are not always valuable to rely on. Unlike hard science, you don't have to prove their significance through statistical analysis

My point is that those metrics are not always valuable to rely on. Unlike hard science, you don't have to prove their significance through statistical analysis, but you should be able to state why market cap/employee is a reasonable valuation tool. IMO, Facebook would not have paid less for 15 employees or more for 100 employees so it seems like a poor metric.