Yeah, I did confuse the discourse there...

My problem with counting cases is that it's heavily dependent on how much testing is done, while deaths are always count.