The SOSP community used to publish theoretically founded OS papers before the 90s. What happened after that? More recent papers, especially after the 1990s, are empirically-backed. They have extensive experimentation to prove that a design works. What is the reason behind the change in the outlook?
I am not saying that previously there was no empirically-backed papers, or currently, there are no theoretically proved design papers. However, the proportion is an indication of the the change in the direction.