It is my observation that many accidents are not reported.
I have friends in the UK who have shared accident cover up incidents with me so it appears to me based on my limited research; your data may be flawed.
As you know, I have always shied away from any formal statistical analysis of non-fatals, because:
a) they're vastly under-reported on the global scale, and no-one can vouch that any particular country is "perfect" either.
b) there are too many of them, and its too big a task for an individual to undertake.
c) we know very well the phases and causes:- take-off, take-off, take-off, landing, take-off, take-off, take-off, landing, taxi... etc...
FWIW, I have just researched my own country again, and find (for AutoGyro/Rotorsport and Magni)
52 reports, plus 1 outstanding fatal, plus 10 "record-only" investigations (only from mid-2019 to late-2021)
A total of 63
of which 4 were fatal (3 if you discount the bizarre civilian ground incident in 2009)
out of a current UK fleet-size of 235.
So more like a 20-1 non-fatal to fatal ratio.
Nothing like 2-1 which someone mentioned up thread.
I cannot comment on what this "means", or how it compares to other countries and the global picture. The comparable data just doesn't exist.
Save that the UK probably has better reporting standards than many other places.