Thats deaths against cases. The true way to do it is at the end of the outbreak.
If you take deaths/(deaths+recovered) you get a much more grim prospect.
And remember the numbers we can run them against are the reported cases. In China a case= hospitalized and tested. We have no idea.
Also deaths are only hospitalized +tested + dead in hospital. If grandma died at home, its just pneumonia.
So, we just dont know yet. The out of China numbers are good, but its only been out in any numbers for less than 2 weeks.
bingo!