The Harris-Benedict study is from 1919 on a population of 239 subjects . I'm just wondering - Do you know if it has been re-evaluated in more recent years, and was found valid?
Last time I looked into it was a years back when I had access to a university library. I dug up a couple of textbooks on sports medicine, they all generally agreed that Harris-Benedict was a good first order approximation.
In practice, it's good enough for a fat guy trying to get thin. It might not be good enough for a heavyweight trying to get down to cruiserweight without losing strength.