Relative to where the government was at the time of the first lockdown in March, it seems much less enthusiastic about 1) doing what is necessary to stop the virus spreading and 2) providing compensation for those affected and to encourage them to comply with new meaures.

This is perplexing. The opposite should be the case. Relative to March, we are in some ways in a much better position to accomplish this.

To begin with, even at this late stage, things are not as bad as they had been allowed to get by the 23 March. It therefore would not take such draconian measures, or for so long, to get control again.

Second, the prospects for a vaccine and in short order are much much brighter than they were at the point we embarked on the first lockdown. This means that social distancing and fiscal support can proceed with more certainty that there is an end in sight.

Third, much knowledge has been gained by other countries' experience of test trace isolate, and also by our own unfortunate experiments. The upside of the disaster is that it doesn't have to repeat itself, and we have good precedents to copy now.

Fourth, a dramatic success has been the spread of remote working, something we were by no means sure could work when the first lockdown began. This should tilt the calculation a great deal.

News about the seriousness of the disease itself is mixed; therapies and knowledge of how to treat critical patients may have cut mortality a lot; but there is an emerging issue of long-term symptoms for a significant chunk of those who don't die.

The working assumption is still that immunity for those who have been infected would not be long term. This should put improvements in the one-time mortality rate into perspective.

As @toxvaerd1 and @giannitsarou point out in one of their papers, waning immunity acts very like a multiplying up of the one-time mortality rate in the cost benefit analysis of a lockdown.