Einstein was a patent clerk. I am certainly not equipped to weigh in on the string theory debates but the ability to make serious intellectual contributions and the skills of navigating siloed academic hierarchies are not necessarily correlated.
It's worth checking out Peter Woit's homepage at http://www.math.columbia.edu/~woit/ and looking beyond the blog and his role as a string theory skeptic. He teaches a number of classes and has a book about quantum mechanics and representation theory that has gotten a lot of favorable reviews. Not sure I'd classify this guy as your average helpdesk guy ;)
I inferred doing the computer support was something he just liked to do. Everyone needs a hobby :)
That's a good point, and probably what GP was hinting at. Thanks for pointing that out :)
He's not an average help-desk guy, but he's not a first rate researcher either. He produced some good lattice qft work in the late 70s/early 80s, then worked unsuccessfully on Chern-Simons theory for a while. Since then he hasn't produced any papers. For a PhD researcher, this is basically 'failure to launch'. (And to be fair, this is probably the most common outcome for STEM PhDs.)