I'm enjoying all the responses along the lines of "He may not have called them tempo runs, but you'd better believe he was running them. When he was well rested for an easy run... TEMPO. When the second half of his long run was faster than his first... TEMPO. When he was racing a 10K... TEMPO."
I'm not saying he wasn't spending some time at threshold one way or another. But when the definition of tempo expands to "anytime he wasn't literally sprinting or walking," the concept kind of loses its meaning.
And for what it's worth, the idea that some runners might not do tempo runs isn't THAT crazy. In fact, it's the entire basis of Stephen Seiler's research on polarized training. The observation is that elite endurance athletes across a variety of sports tend to do (simplifying here) about 80 percent of their training very easy and 20 percent very hard, with not much in the middle. There's some data on polarized training in runners, although the picture isn't as clear as some make it out to be.
I can certainly attest that when I was in college in the 1990s, our team did zero tempo runs (and that includes zero "accidental" ones). I'd read about them, and started sneaking out to do some 20-minute Daniels-style tempos on my own when I was a senior, and it was a completely foreign stimulus.