Weekly Covid log graph update and the upcoming data correction

I added the data point for the week that ended on 15 July. A 7-day rolling mean of 77 kept my line straight. Yesterday it was announced that new data was being suspended because of a controversy over what was being reported. Today I read in The Times that the deaths that might be removed from the statistics might account for 10% so I multiplied the numbers by 0.9 and replotted the graph - orange line above. Of course the gradient remains unchanged. If those deaths represent a fixed fraction then the numbers will not come down at a faster rate; they will just reach the bottom sooner because they started from lower. Since the issue is about people who had Covid and died over 28 days later, it won't quite be a constant factor though. More should have entered the statistics as time has gone by. This might change the gradient part way through.