I have exactly zero faith in any estimates at this point because no one really knows where this is going to go.
Make your own rough prediction of the next few days.
Go to the JHU dashboard, select US, select "Daily Increase" in the bottom right. You'll see that the daily increase itself increased pretty much every day.
Now select several other countries, ideally ones further along in flattening the curve, and see how their daily increase behaved relative to the peak daily increase. For example, Germany's daily increase stopped increasing about six days ago and is now slowly decreasing. South Korea took about ten days from the stop-of-increasing-increases to decrease down to manageable levels.
Armed with that "cohort of countries" behavioral data, ask yourself two questions: First, is the US at the peak daily increase now, or is it going to keep increasing? (hard to answer, I personally doubt it) Second, once peak daily increase has been reached, is the US faster or slower in getting that increase down again than the countries you compare it to? (I'm guessing a lot slower than South Korea, a bit slower than Germany - mostly because not all states/counties/whatever adhere strictly to flattening the curve)
Using those two answers you can roughly predict the future of the US daily increase.
Similarly, you can use the US confirmed case count, confirmed case age, and confirmed death toll to estimate how many unconfirmed cases there likely are (a lot) and how many yet unresolved future deaths there will be (a lot - recent infections don't have their outcomes yet, and the biggest number of recent detections is recent [not quite the same as recent infections, but it's the best data we have]).
Theoretical best case scenario, the US is already at the peak daily increase and does a South Korea right now, ie a few more days of those peak daily increases and then steep decreases... based on SK numbers that would still be 200-300k additional cases during the two weeks of decreases, and 1k daily from then on. Sadly, I don't think that best case scenario is realistic.
Another good case scenario, the US is already at the peak and does an Italy right now, ie continuing at constant daily increases - that's 26.4k daily new cases, or 185k a week.
A less good case scenario, the US keeps increasing their daily increase for a week, and then does an Italy. During the last week, daily new cases went up 2.5x, resulting in a steady state of 66k daily new cases. If 10% of those need an ICU bed for two weeks that's roughly all US ICU beds constantly occupied by covid patients
[src]
Personally, I'm guessing the US will be somewhere between scenario 2 and 3... I'll try to remember to come back at this prediction in the future
For illustration of my pessimism, US daily increases as of now:
Nothing in that chart suggests it's getting better right now.