That NY antibody test mentioned above does give us some interesting data that we can use to extrapolate actual cases. Their findings are that there are 2.7 million people infected in NY, an approximate 10 fold increase on the official number of 287,245, which give a case fatality rate of about .7% (instead of 7.3%), which is in line with the estimated range of 0.6-1.2.
If we take that CFR and apply it to other countries it should give us a reasonable estimate of actual cases. So for the UK, the FT is now saying there have been about 42,000 excess deaths, which gives us a range of about 3.5 - 6 million actual cases in the UK rather than the current total of around 150,000.
In Ireland at around 1000 deaths it means an actual infection rate of 70-12000 instead of the official number of 20,000.
Sounds crazy but the NY study seems fairly reliable in terms of methodology and sample size and it means that the vast majority of cases were either mild or asymptomatic - which is good news.