If you compare granularity of tracking and utility of historical version control across different projects of that era, such as FreeBSD vs. Linux, I think you'll find that Linus' position is clearly false.