re The v low deaths rates atm:
In the early days and at peak infections / deaths, we were all massively under-counting cases compared to now,
when we are testing more and targeting those tests too.
I've been looking at the attached useful table today for Germany's test figures over the weeks
(tests not people, btw and to make it even more complicated, over the last week there were another 80,000 tests at points of entry which may not yet be fully included in the test totals !)
Anyway, positivity at peak COVID with > 30,000 weekly cases was about 9 x higher than at present
So maybe we should be comparing deaths and hospitalisations for very roughly [270,000 = 30,000 x 9] cases at a peak week vs the 8,400 cases last week
(and then multiply both by 2-3 to get "true # cases)
i.e. by my rough estimation, at peak Germany had about 30 x the current # cases
i.e. lockdown and social distancing have currently reduced cases x 30
The UK would have a similar reduction, probably even more due to its v long lockdown
That and having finally sorted out testing, track & trace is why GB is finally showing the expected epidemiological benefit of being an island - that we see in other islands - and having far less of a rise in cases than its neighbours
Deaths in Germany for many weeks have been in single figures, say average 5 compared to peak of about 350
Then consider that the average age of infection has dropped about 12 years since peak
==> the current single figure deaths look very understandable
==> the low deaths around UK / Europe make sense
imo, even if the age of infection increases a lot over winter, no European country should return to anything like the peak deaths, providing we don't return to anything like the (real) peak cases