That modeling has been out for maybe a week now, maybe a bit longer. When I first saw it I was really hopeful. I don't know though. Just anecdotal evidence makes me question how right they are.

For example, I read yesterday about a church choir in Washington that wasn't sure whether they should have choir practice three weeks ago. 120 members. They decided to go ahead and have it but only 60 of them showed up.

Now 2 of those 60 are dead and 3 are/were hospitalized (not sure if that means 3 total were hospitalized and 2 of the 3 died or if that means 5 were hospitalized and 2 of the 5 died so far). 45 of them tested positive for corona virus.

That's not a randomized and statistically valid trial but in terms of anecdotal evidence, it makes it seem somewhat unlikely that the infection rate is as much higher than the the hospitalization and mortality rate as what the Oxford modeling suggests.