Magnitudes greater than 5000 people though for any decent degree of confidence. Equally changes in care protocols as time progresses and we understand the disease more may change the figure a fair bit...

I think a bit lower down from your comment there's a post from nk15 which shows the difference between South Korea, Wuhan Princess and Chinese data samples on death rates. This shows you how diverse the death rate range can be- from 0.7% to 2.3% dependent on where it was caught. This illustrates the difficulty in estimating at the current stage.