What I have read about Italy in the last few days is that the Lombardy region is improving (actual lower daily cases, lower daily deaths) because of their early lock-down, while the rest of Italy is still on the standard exponential curve.
Italy is approximately 8 days ahead of the USA. Infection to resolution is 10-14 days. So, any true peak in mortality simply indicates the primary wave of infection occurred two weeks previous. So, if you really want to have an excellent bird's eye view that is 98+-% accurate, and is only subject to modeling assumptions going forward, then watch Italy, US and NY.
On that note, look how NY has distorted total US deaths due to their extreme growth curve over the last week. Since NY is a lagging indicator, this is demonstrating - to a very fine level of granularity - how compressed CV is in dense urban regions, not an indicator of potential US trends:
USA NY Day Date Cum Total Cum Total NY % 1 2/29/2020 1 2 3/1/2020 2 3 3/2/2020 6 4 3/3/2020 9 5 3/4/2020 12 6 3/5/2020 14 7 3/6/2020 18 8 3/7/2020 19 9 3/8/2020 22 10 3/9/2020 26 11 3/10/2020 31 12 3/11/2020 38 13 3/12/2020 42 14 3/13/2020 49 15 3/14/2020 56 2 3.6% 16 3/15/2020 62 3 4.8% 17 3/16/2020 75 7 9.3% 18 3/17/2020 96 12 12.5% 19 3/18/2020 122 16 13.1% 20 3/19/2020 174 38 21.8% 21 3/20/2020 229 46 20.1% 22 3/21/2020 296 58 19.6% 23 3/22/2020 405 76 18.8% 24 3/23/2020 516 157 30.4% 25 3/24/2020 670 271 40.4%