You're right, and I don't doubt the value of studying open source projects or that it's the best that can be done given the data available, but there are limita

You're right, and I don't doubt the value of studying open source projects or that it's the best that can be done given the data available, but there are limitations on how much light this data can shed on the great mass of code being written within organizations. Maybe the results do apply to the situation you describe, I'd be curious to know that.