This is kinda where I'm at right now. The true mortality rate is almost certainly well below the 3% being reported now due to a high number of mild unreported cases. That said, the speed with which it is spreading combined with the potential severity is cause for concern.
If we compare this to the previous SARS and MERS events the trajectory is worrying. The SARS epidemic lasted 19 months and resulted in 8400 confirmed cases. The MERS epidemic has been ongoing since 2012 (very few cases after 2015 though) and resulted in 2500 confirmed cases. The current Coronavirus epidemic has only been going for a little over 2 months and we're already over 80,000 confirmed cases with no indication of a slowdown.
As a counterpoint though, from October through December of last year, there were 3.7 million cases of the flu in the US alone.
If you charge $1,400 to each suspected person then either the testing will bankrupt the population or people are going to hide from testing. If coronavirus arrives in bulk in the US, then it's either going to be "Medicare for coughs for everyone" or a pandemic.
> Hospital officials at Jackson told the Miami Herald that, based on his insurance, Azcue would only be responsible for $1,400 of that bill, but Azcue said he heard from his insurer that he would also have to provide additional documentation: three years of medical records to prove that the flu he got didn’t relate to a preexisting condition.
Also for the irony:
> Azcue said he earns about $55,000 a year working for a medical device company that does not offer health insurance
With the deaths in Italy it is now also unclear if one person actually died from COVID-19 or if it was the late stage cancer they had.
Assuming that 20% of the US population (330 million) gets it, we're looking at about 13.2 million (20%x20%x330 million) patients who require admission to an ICU ward. Needless to say, that is enough to overwhelm the healthcare system and hospitals, and increase the overall mortality rate, when people who requires ICU are turned away due to lack of beds or equipment.
As it is, the best option would be to delay the spread as much as possible, so that the number would be spread across a time-span for as long as possible.