bit more from the article and comments
"...the researchers suggest that the vast majority of people who contract Covid-19 may suffer little or no illness. They believe that as few as one in a thousand could require hospital treatment — an assumption that suggests, according to their model, that half of the country could now be immune...Even if it’s one in a hundred [who fall seriously ill] you still get to 35 per cent immunity,” Sunetra Gupta, a professor of theoretical epidemiology at Oxford, who led the study, said.
Jonathan Ball, professor of molecular virology at the University of Nottingham, urged that serological tests begin but said: “This is interesting work, but hampered by the same issues that impact all epidemiological models — they rely on assumptions based on a paucity of fact about how this virus transmits.”
The findings are based on assumptions about the most likely characteristics of the disease and are yet to be peer-reviewed or published in a journal."
Comment - Extremely unlikely that 50% of the population are already infected. We can’t disprove this of course but if we take the lowest case fatality rate of 1 in 1000 (from figures from areas all over the world) then we could have 30000 deaths - a disastrous tsunami of cases and deaths facing the NHS and the nation in a matter of weeks. The more realistic assessment is the UCL prediction that 10% of London’s population could/would be infected.