The vertical coordinates for the chart above are computed by dividing the number of confirmed reported COVID-19 infections by the number of reported infections reported on the first day that there were more than 100 for that locality. The horizontal scale is the number of days the reporting date is after the first date of 100 or more reported. All locations start at 0 days and 100%.
This chart shows the dramatic flattening of growth that both Hubei and South Korea experienced due to their government's actions. It will take close to a week to see the impact that this weekend's shelter-in-place orders will have. The average incubation period for the virus is 5 days. I expect to see continued growth in the US as the testing backlog is addressed. Whether it is resolved fully or partially, in either case the number of confirmed infections will continue to grow over the next several days.
I cannot predict when or at what level the flattening of the curves will be because there are many variables including relative testing adequacy, population compliance, and population density. South Korea has demonstrated the power of an aggressive testing program to identify and isolate carriers of the virus and their contacts. The struggles of the US testing program have been well covered elsewhere but progress is being made.
New York City is one of the most densely populated places in the world. The ability to achieve social distancing will be a large challenge. The relatively high dependence on things like mass transit make containment more difficult. We have seen the stories of the non-compliance in Italy that has allowed its curve to hardly bend. Florida's allowing of bars and restaurants to stay open may not affect their statistics significantly because many of the spring break participants are from out of state. They will go home and infect their families and possibly create another wave of infections.
It should be noted that the data for Hubei, China began on January 20, 2020 and in actuality would be moved higher by about 30% of the distance (doubling in nominal terms) between each grid line and to the right by 3-5 days depending on how long the authorities knew there were 100 or more cases before the WHO reporting began. The shape of the line would be the same except in the lower left hand corner where the additional data points would have been.