If I "touch a.txt" then run "ls", and see "a.txt", I'd consider my filesystem realtime. There's no point in being pedantic about "but there's a 50ms delay when I hit enter after typing ls so it's not realtime!!!".
"Realtime virus outbreak" implies that the site is somehow automatically identifying where the virus is and updating it's stats, rather than waiting for a human to update it.
What would you recommend calling it instead?
In the given example there's a tracker (updated by humans) and you can see the current state of the tracker at any time (from their database).
I'm not sure what the distinction is here if you consider one realtime and the other not.
Real-time is not all about latency, live TV or a Skype call also has a latency but it is real-time for all means and purposes since it is automatically gathered, transformed and transported without human labour and immediately.
Almost all of it is done by hand by the folks at the Johns Hopkins University's CSSE department. It gets updated around twice a day and they're constantly fiddling with the layout and whatnot.
No, it's updated quite irregularly. Given the time difference between the East Coast and China I'd expect that quite a bit of delay is to be expected between information being released and CSSE updating their site. They're also combining data from different sources and clearly are batching updates together.
About the update rate: It's updated regularly. If you want to go down this rabbit hole, is 1m update real-time? Is 1s update realtime? Where do you draw the line? For most people, anything less than 12h, in the case of tracking confirmed cases, is perfectly fine described as "real-time"
About the data: Obviously it's showing confirmed cases and not actual health of people. The UI makes that pretty clear, it says "Confirmed cases" in big font. No one expects the latter...