What is the industry standard for number of servers per Unix/Linux engineer and are there any white papers or the like that discuss this?

There is no standard since it depends critically on what the servers do, how big/complicated they are, how much the environment is changing, how critical the servers are to the company, how much downtime you can tolerate, etc.

I've worked for a small company where we had 3 admins for under 30 servers - but it was an internet company where things were constantly changing and apps provided on those servers were the entire revenue stream so we had to have system support 24x7. I've also worked at a large company where most of the servers were linux blades arranged in huge farms of identical machines where we had something like 200 servers per admin and it worked fine since so much of the admin work could be automated.