Their other papers are about BigTable or load balancing systems that the researchers developed for an immediate business need of the company, and then published

Their other papers are about BigTable or load balancing systems that the researchers developed for an immediate business need of the company, and then published an overview of their work incidentally after the fact. I have seen very little come out that was even analysis that wasn't necessary to do internally if the company was never going to publish anything. Maybe I'm wrong, but it seems like Google and Amazon follow this model much more than other companies like IBM or Microsoft who more often publish much more theoretical on not-yet-practical work that is only in the direction of their goals rather than already productionized well before it gets published.