First let me say I don't have a rMB - but I do use R and Stata on a MBP and I'm currently looking for a second hand rMB as a travel machine. Anyhow, its not just about the data, but also about the kinds of models you will build. You could have a large dataset and a simple model will still converge quickly, or you could have a fairly small dataset and set up a complicated hierarchicial Bayesian MCMC model that takes days to converge. So there is not a simple answer to your question - it simply depends what you are going to do on the machine. That said, if you anticipate doing anything power-hungry you would probably be better off with an MBP. Although I'm looking for a rMB - I mostly just plan to use it for writing, emailing etc when travelling, maybe but beta OS versions on it etc. I will put R on it but would not anticipate doing anything taxing on it.
