This looks like false conclusions from oversimplifying the death ratio above health care saturation.

You can't just change the axis labels because the death ratio depends non-linearly on infection ratio above the threshold.

Are we willing to take the risk when we don't know the consequences?