so, is the data in italy beginning to plateau, indicating a so-called flattening of the curve? it's a bit choppy, on first glance, but given the inevitability of deficits in reporting accuracy, it's plateau-y enough, it would appear.
i have two reasons for you to read that data skeptically.
the first is that there are reports of a large amount of uncounted deaths, particularly at nursing homes, during the peak of the crisis a few weeks ago. the delay in counting these deaths might have muted an actual peak. the second is that the actual situation on the ground may be a series of localized peaks, which may appear as a plateau if plotted on top of each other in a way that obscures the overlapping peaks. that might actually explain the choppiness of the data - it's a series of maxima, not just one.
it would be useful to see a couple of different test cases in order to help work through some of these ambiguities in how to interpret the data from italy. i believe that the same concerns exist for spanish data, as well.
i think that the data from north america is likely to continue to be presented in very regionalized contexts, mostly by states and provinces, but also by mayors of large or important jurisdictions. hopefully, the more regionalized presentation of the data will help clarify the second point - it should make it easier to identify whether we're seeing spikes or plateaus, which is a marker to determine if we're suppressing the virus or if it's burning out.
as to the first, it is possible that delays in counting the dead might be better separated from the curve, if it is more localized, as well. but, north american jurisdictions seem to be more aware now, eventually, of the virus' particularly deadly potential inside of these care facilities, and are finally moving to take serious steps to protect these people from this. so, there is also a better likelihood that we'll actually get correct statistics from these homes, when the peak or plateau actually arrives - we might succeed in eliminating this bias by being prepared for it.