R seems much more easier for quick statistic, modeling, and rapid prototyping.
Python would be better in general minus some statistical libraries and loses some ease of built-in stat function that R have.
I could see myself investing much more in R for my statistical career if I were mainly a statisticians that does data cleaning and data modeling.
Python would be more well rounded if I were in a position that is more computer science/programming. Or if I want more speed I would move my code from R to Python.
BTW, I'm going to choose R for now for my statistic graduate major, there are many R books and stats books that uses R.