> The differences in performance between these cars is so small that in fact the driver makes a lot of difference in a relative sense and not an absolute sense.

This is a really good way of putting it.

Is it? How do you end up with dominant drivers for years that magically become non dominant if the difference is “small”. See Vettel, Hamilton..

I’d say the difference is large.

You're discounting the human variable there. We're not machines. Look at other sports and you can see other athletes be dominant for years, then "magically" become non dominant after a single bad season (sometimes after a single bad game!). Factors like age and confidence play a huge role in human performance.

I don’t disagree. But in these cases I’m more inclined to believe teams bridged large engineering gaps between competitors or teams failed in adapting to new rules. My core argument is that if the engineering deficit was small, domination would be a rarer trait in F1