Maybe things have changed since I last programmed seriously in R nearly 10 years ago, but back then the language had serious design flaws that made it way too easy to write incorrect programs. Conflating scalar and vector types, weak typing, and various other design choices led to hard to fix bugs, even you even realized the answers were incorrect. Proponents of strong static typing might say the same sorts of things about Python, but Python is much better than R in this regard.

I would hope that data journalism, like any other kind of journalism, would care about correctness.