[timcoote (OP)]

We heard on the Cambridge Conversations how we do not understand the immune response to the virus, nor do we know whether a vaccine can be created.

Presumably, as it's a 'whole population' infection, with quite varied impacts by age (qv David Spiegelhalter: "How much ‘normal’ risk does Covid represent?"), getting the balance of risk/effectiveness for a vaccine is tricky and time consuming.

So, getting an early estimate of the likely immune response and duration at the population level would be a useful tool for the scenario planning for different interventions, especially as it seems that the demand on ICU beds is quite low for younger people.  Is it practical to estimate such an effect with confidence intervals, so that we can get a view on the value of the option of different lockdown exits? 

The costs of the exit approaches outlined in Ken's slide of Ferguson and Harvard's lockdown models look very large.