Depends on the leisure though, my impression is that road enthusiasts are relatively higher risk (a factor of both attitude and that rural single carriageways are higher risk) so that can work both ways.
I'd agree that it would be rash to draw conclusions, I was being somewhat flippant in my statement partly as the initial reporting was poor. On top of the behavioural oddities of 2020, I suspect there is also significant uncertainty over the distance estimate that is unlikely to ever be resolved. Cycle counts are limited and patchy and the NTS does not include circular journeys that have no purpose other than the journey itself.
Danger is the possibility of harm etc, there are absolutely ways to objectively measure that though when it comes to transport I'm open to quibbling and/or context on whether the rate part of casualty rates should be time or distance.
Of course, perceived danger is a related but separate issue and is entirely subjective. We already know that most people vastly overestimate the dangers of cycling in relation to many other activities (hence the h-word debate) and I doubt that sensationalised headlines and low quality statistical journalism as demonstrated by the guardian above are much help with that.
It's also perceived danger which is the relevant one when it comes to peoples' transport choices and why the dutch approach works and the vehicularist one does not.