I don't feel like this answers the last part of my question: why would that be a useful metric? If the claim is that I should be concerned with the future of kernel development because of the future of Redhat, why should dollar amounts matter rather than directly inferring from proportion of the development they do independent of dollars?