17% of 3.5M is 600K. Number of cases = 26,000
that is more than 20X the number of uncounted (but not undetected - people have been told to self-isolate at home and are than never tested, never counted) cases.
The "death rate" has gone down, from a CFR of 20% to the actual 0.9%.
Not good news, since no-one takes CFR to be death rate, except you? Just expected news. There is no reason to expect large numbers of non-antibody positive infections. There is no reason to think the remaining 83% will not have broadly similar IFR if/when they catch it in a second wave.
0.9% is a bit high for IFR (previous estimates 0.66%).
Inner demographics is actually quite young, so does not explain this:
London’s population is young (average age 36.5) compared to the UK overall (40.3). More than one in 10 people living in Inner London (11.7%) are aged between 30 and 34. This compares to just 6.2% of those in the rest of England. More broadly, in Inner London, almost half the population is made up out of those who are in their early twenties to early forties (47%), compared to the rest of England where three in 10 (31%) are in this age group, and Inner London is home to a higher proportion of young people than Outer London.
I gave possible things that might reduce this a bit, but it could be pretty accurate. That would not be good news, and more than 50% higher than previous expectations, taking demographics into account.