If it hits a true 3.4%, that's a pretty big deal. I doubt that number, because they're only running off known cases.
The death rate numbers now are wrong, and almost certainly understated, for a mathematical reason. The reason is that the virus is still spreading. When spreading, it means each infected person infects more than one other person...by definition, otherwise it would be contained. For a hypothetical Budweiser virus with R=2, that means 10 cases today, 20 tomorrow, 40 the next, etc. Meanwhile it takes a couple weeks to die from it. So by the time the first 10 die, there are 100 people infected, and they say it has a 10% death rate, even if it's actually 100%. Because incubation periods and when the virus was contracted are never well known, there is no practical way to correct the statistics until the virus burns out. Which hasn't happened yet, except probably in some isolated populations in China. So you can regard all death rate estimates as BS, not because they are lying, but because it's impossible to know yet.