Example, a user name might be "ekg7d9a", but the user's real name is "Charlie Hanes", and the email is "chanes@example.com".
What are the dominant reasons to use non-related user names with logins and email addresses vs an at-least-equally-common method of having "Charlie Hanes" using "chanes" as his login?
An approach which can indeed be useful is using the company employee ID as his/her login (in companies where employees are assigned an ID, of course); this allows for easier mapping between AD and other HR applications.
Also, this removes the headache of handling duplicate names, which can become a real issue when lots of people are involved in the company.
The problem we run in to a lot here is people insisting on changing their usernames due to marriage/divorce, etc. Usernames not tied to a demographic piece of information completely sidesteps the entire issue.
Because employees are in fact "human resources" and can be identified and tracked by using non-personally identifiable identifiers just as you would with any other business resource.
Security's an oft-cited reason (though one of dubious value, in my opinion). Particularly for high access individuals, there is some merit to not linking their user ID to a publicly available piece of information, like their name - it's another piece of information an attacker would have to acquire before getting access to a given individual's account.
In theory it makes some sense, but in my experience, no one's network and devices are locked down hard enough to make it actually difficult to acquire a person's username, with even very minimal "access"/visibility.
