I think you did a good job on the data presentation. I'm not even a vim user and I found it intuitive to understand the analysis you did (i.e. the question you wanted to ask) and the results of that analysis. Just generally speaking, I thought it was a very data science-esk solution. You had a curiosity (what settings do people use in their default vim config), you came up with a way to get data that could help shed light on that question (grab vim config files from a location that has a lot of people from a relevant sample (developers) revealing their vim config files -- github -- which you use as observations, even though the people who originally uploaded those files had no intentions on them being observations in some person's study), and then you presented various analyses of that data in a concise way.

It may be a simple and trivial question to some, but more importantly, I think the methodology was very well done because I feel can trust your results, which should be the goal of ANY empirical analysis.

Yeah, fun idea. So when I didn't recognize some of the options, I've looked them up with :help and of course "autoindent" and "ai" are the same, so should be really added together.

Will you release the source? I don't use any of the stuff you use, except vim, would be interested to have a look