The USA may have peaked in deaths per day from COVID-19 (sometime around April 16-19)

The USA was supposed to reach a peak in mid-to-late April in terms of the COVID-19 pandemic caused by the SARS-CoV-2. I have not been paying attention much to the "new cases" or "total cases" since those numbers in the USA reflect how many people are being tested and what criteria are being applied to decide who will be tested, and the results may not reflect the actual spread of the virus very well.  I have instead been paying attention to deaths attributed to COVID-19.  That said, even the data on that are highly suspect; many persons die without being tested, and it is rare to waste a precious test on a cadaver. When epidemiologists compare deaths this year to deaths a year ago, any increases probably show the influenced of COVID-19. The non-pharmaceutical interventions (almost everyone staying home) reduces communicable diseases such as the seasonal flu and deaths related to people being out and about (drunken homicides related to fights at bars, traffic deaths, etc.), so keep in mind that deaths-per-day related to COVID-19 are still rough approximations of the problem. I've observed four sources for estimates of deaths per day: the World-o-Meter Covid-19 update seems pretty good. Johns Hopkins University is another good source. Our World in Data and CNN also have some good maps (but their color schemes don't offer enough contrast) and data sources.

Here is a chart from Johns Hopkins showing deaths per day in the USA:

I downloaded data from NBC, put them in a spreadsheet, and ran the death rate with five-day averages (every day's death count was replaced with the average of the previous two days, that day, and the following two days) to smooth out the chart, and came up with this:

The data sources are different, so the charts look a little different, but both charts show a decline just now (April 19th and 20th seem to have had fewer deaths than the preceding several days).  This may be a sign that we have actually peaked in our deaths-per-day, or it could be something like the dip in deaths-per-day that the NBC data have for April 10th-12th.  

Deaths seem to come (on average) about 18 days after people get infected, but the range is huge (something like 15-45 days from infection). Because the length of time between infection and death isn't normally distributed (it is very skewed; almost no one dies within 10 days of infection, and most die in the 12-24 days after infection, but still a lot of people die in the 25-50th day after infection), we should expect the decline in deaths per day to be much more gradual than the steep increase we saw in March and April up to the 16th of April. 

The orders to stay at home and the way people have been obeying those orders, as well as the habit of wearing masks in public have seemingly helped us as a country stop the growth in deaths per day.  It is important to wear masks in public because a quarter to three-quarters of persons infected by the virus are not showing symptoms because the virus is slow to express itself in symptoms and many persons (currently plausible estimates made by experts and researchers vary from 20% to 80%) never do show symptoms never do show symptoms, but asymptomatic and pre-symptomatic persons can spread the virus. Wearing a mask reduces the water droplets that float in the air around you as you breathe or speak, and thus reduce the virus count you are putting out (if you are infected) into the air around you.  Inhaling a billion viruses is likely to cause a worse problem than inhaling a thousand viruses, so please wear the masks to show your solidarity with everyone and show your recognition of the fact that it is easy to be a carrier of SARS-CoV-19 without knowing it.