The data is flawed. Countries use test different. In Sweden, for example, there have not been any extensive testing and there won't be either. So the numbers do not reflect how many that has had the virus. It's a tip of the iceberg.
In the beginning the swedish number could more or less be stated as people that have been so sick that they had to go to the hospital. Then they got tested. People that have had mild symptoms have been ordered to stay home in quarantine and we have not been tested. I had to stay home for more than 2,5 week and if it was corona, a cold or mild influenza, I dont really know.
Now a little later some personal on hospitals has been tested but as stated. They say there won't be any masstesting on the general population.
Yesterday they mentioned that they had done a randomised study to find out how many that was sick in the Stockholm area (1-2million people depending city or region). So about 773 people was tested and 2,5% had the virus at that time of testing. That was more than a week ago. Now they think 5-10% might have it actively. There will be some models published later today and I hope it also includes numbers regarding people who did have the virus but do not have it actively. I can link to it later if you want.
Anyway, that random test show that alot more than ~9000 has been infected. And let me include this.
Next week a similar study will be done across the whole country. I think they said 4000 people was suppose to be tested.