I'm pretty sure the figures arrived at by the WHO, CFR of 2%, are based on the assumption that there are many people who have carried on unaffected without presenting to the hospital amd4578
Of you were just looking at the hospitalised patients you get a much bigger CFR risk than 2%. On worldometers it says:
"A study on 138 hospitalized patients with 2019-nCoV infection, published on February 7 on JAMA, found that 26% of patients required admission to the intensive care unit (ICU) and 4.3% died, but a number of patients were still hospitalized at the time. [9]"
(So, I guess that study isn't holding out much hope for those who are currently critically ill and expects this death rate % to climb)