but my point is that cases identified isn't the right denominator as if you aren't a recovery or a death, you are an unknown.

Calculated that way in the US, there are 31 deaths and 15 recoveries. (66%) I don't think that tells us that the method is bogus, just the sample size.

Switzerland is at 57% (4 deaths, 3 recoveries); same thing, ridiculous sample size.

Iran is 354 deaths and 2959 recoveries (10.5%)

Germany is 3 deaths and 25 recoveries ( 10.7%)

China is at 3,136 deaths and 59,913 recoveries (4.8%)

Japan is 19 deaths, 459 recoveries ( 4.0%)

Italy is at 631 deaths and 1004 recoveries (3.9%)

Spain is at 49 deaths and 183 recoveries (2%)