It's a ceteris parabis statement. So other things such as suppleness and tread (knobbies) matter too. When you're comparing two different bikes, presumably things like fit and aerodynamics matter too.

The person who did a bunch of research on the topic, Jan Heine, suggests that "wider is faster" is only true up to a point. From what I recall, it was based on experimentation rather than a theoretical reason. I don't recall a satisfying explanation from him. I remember wider is faster being true to something like 30-35 mm. But it's a fuzzy memory.