"...deaths numbers would accumulate before the recovered numbers and amplify the perceived death rate..."

It is good you note this, because that is very true. It takes longer to survive than it does to die, and this skews stats.

Presently, due to the chaotic conditions of the epidemic, it is impossible to derive accurate figures for any metric regarding nCoV in China. The data provided by official CCP sanctioned sources is impossibly smooth and appears to be a deliberate attempt to project the ~2% mortality rate.

Thanks!