it seems like a much shorter, shitter and weaselier version of this article
www.theatlantic.com/health/archive/2020/08/how-to-test-every-american-for-covid-19-every-day/615217/
the graph is a fine piece of fraud that someone from the centre for (supposedly) evidence-based medicine. One line shows 'new cases'. Another shows the number of patients in hospital, which is not the same thing at all.
The start date on the x axis is April, but it's my understanding that the data are not comparable before 2 July at the very earliest due to the aggregation of Pier 1 & 2 data.
It was obvious months and months ago that the age of coronavirus patients was massively higher than the average age of infection, because young people were often asymptomatic.
We should expect a younger population once we start testing people millions of people monthly rather than just a few at death's door.
Also it does seem that a lot of young people have concluded that covid-19 is not necessarily a major existential threat for them, and are going about normal daily life, so we should expect the age to get younger.
It's of course true that the PCR test might not be the best way to test people, but this doesn't necessarily seem to be related to the data presented in the article.