Well, there's a delay between what intervention measures you take, and the effect on the numbers. If we can keep the curve flat, then there won't be a surge in cases which could overload acute bed capacity. If it comes back to zero, then we risk a rebound if restrictions are relaxed to much. It's too late for eradication, so we need to keep the disease spread slow enough that the system can cope with it. I suspect that if the numbers drop off steeply, they'll relax restrictions a bit to keep a steady stream of new cases, with no sudden peaks. The same amount of people will become infected, the same percentage of those cases will need hospitalisation, and the same smaller percentage will need ICU support. If we keep the numbers low enough so that the numbers needing ICU are less than the ICU capacity, then slightly less people will die.