I think you're right. The issue is, due to the power law distribution of talent, the sample size of such roles (and personalities, i.e., how many Python BDFL's are there?) is incredibly small, so it makes precise assessments difficult and the variance increases wildly since there aren't very good comparables.

I think he could command nearly $10 megabucks, give half of it away to charities, grants, scholarships, social ventures, and underserved areas to promote STEM college prep, and still have more money that one ostensibly average, middle-class person would know how to invest (real estate on the near periphery of burgeoning urban sprawl areas). Microsoft is a cash cow and is willing to invest in a solid, deep bench to stay competitive, especially in an apparent transition to either EEE or adopt OSS.