I found James Ward's models pretty meaningless as he did not provide details of any of the underlying assumptions.
He says initially if we delay reopening we will have no cases through the winter but then have a larger wave in the spring. This is the premise of his arguement throughout and from what I understand that of sage.
However two charts down he goes on to say that 'I am very nervous about using the model to predict what is happens in the winter or next spring there are so many factors here that are unmodelled or uncertain'.
So basically he is willing to risk more deaths in the short term for an unknown outcome in the long term.
He the notes he is not qualified to comment on the risk of new variants which to me is one of the most fundamental reasons to keep numbers down as much as possible. There is also no consideration of the impact of the arrival of the beta variant for which AZ has only 10% immunity and is likely to increase as antibodies to delta increase and delta declines.
I have done some modelling as part of my work (in a completely different field) and you can develop a really sophicated model but its accuacy basically comes down to the underlying assumptions and data, if these are incorrect the model will be completely wrong. In relation to covid James is entirely correct in that there are way too many variables that are highly uncertain so it is impossible to know how things will turn out in the long term.
On this basis I would come to the opposite conclusion to him - that we should do what we can to reduce deaths in the short term and reduce infections to mitigate the risks of new vaccine resistant variants. We also need to reduce the spread of existing vaccine resistant variants like beta.
We have no idea what the impact of a wave now will be on cases in the winter or spring so why risk deaths now for an unknown?