I will agree that it’s not an easy figure to come up with, after the fact.
But I think 4-6% is an underestimate, not an overestimate.
I’ve been to ER several times for hypos and probably should’ve been brought to an ER a couple other times too. But other than initial diagnosis I’ve never been to hospital for hyperglycemia.
Since way more than 6% of my ER visits have been for hypos I’m very willing to believe that the hypo death rate for T1’s is more than 6%.
Never mind the scary things like the possibility of me wrapping a car around a tree because I’m hypo, killing me AND my family. I mean, the scare is good, it makes sure I check my bg all the time before driving.