>>I had the bad luck of having to re-write his R code into Python which was suitable for this specific problem than R was and had to explain to this junior dev. Still, he wouldn't budge and kept writing untested production software in R when in reality most of his peers already knew Python and Ruby, whereas he was the only one with R knowledge and no one else in the company could understand his code, not even people familiar with R.

Why was he given so much freedom in writing code? Especially as junior? I'm a senior at this point and I couldn't just arbitrarily decide to use R(or any other language that isn't C++ for that matter) for my job - it just wouldn't go through code review. Because I'm sure you had code review at your work, right? And if you did - who was approving his mess of a code?