What should they be using?

If someone understands this crude metric as "If you catch it and you show symptoms and you test positive, your likelihood of dieing is roughly (deaths divided by confirmed cases)" then that seems reasonable to me.

"Try to avoid catching it." is reasonable advice, as is "Be aware that you could catch it and be asymptomatic while still being infectious." to discourage people from visiting their grandparents this month.

I don't feel like I'm being misinformed if I divide deaths by confirmed cases, I feel like I'm getting a rough worst-case estimate.

I've been using this[0] page as a reference, which seems to back up those numbers. It currently shows 82,424 recoveries and 8,697 deaths. Can you recommend a better resource I should be looking at?

[0]https://nssac.bii.virginia.edu/covid-19/dashboard/

It’s not this simple. It takes a longer time for people to recover than it does for people to die. Particularly when you consider that people stop having symptoms, so “recover” in the ordinary sense of the word, but still test positive for the virus, so are not marked as recovered.

Look at the mortality specifically in places which overtest: South Korea, Japan, Taiwan. There it looks closer to 1%. There are additional factors to consider like population age, but there Italy is the outlier, not South Korea.

That would only list those recoveries that have been bad enough to warrant a test in the first place. With health care systems being overwhelmed the world over, almost no country is doing extensive testing of all people who suspect they might have been infected, meaning there will be a huge number of mild cases going unreported. South Korea is perhaps the only exception.

South Korea has done the most extensive testing of any country and there death rate is running around 5%. But here we are playing "I’m not an epidemiologist but".

This is false. Please stop.

Current fatality rate is less than 1% of confirmed cases.

I love how you are basically going around the thread with bold statements with 0 data to backup your claims regarding the fatality rates.

Currently experts are giving us data and people are giving us sources.

[1] Countries like Iran literally have mass graves at this point and here you are claiming this is overblown by social media.

You really need to stop.

[1] https://youtu.be/eWUT1V6WrRA?t=12

I think you're wrong. I explained more here https://news.ycombinator.com/item?id=22622231

Suspected cases