While we did manage to flatten the curve in spring, growth is back on an exponential curve in Canada, which means the threat of overwhelming the health infrastructure is still looming. A sweeping lockdown may not be the best answer, but measures needs to be taken to slow the growth.
Exponential growth is no joke. Hospitals are more prepared, but exponential growth means that even if you double the hospital capacity, they will still be outmatched rapidly if nothing is done to slow down the spreading.
I think you should tell that to the residents of New York City, who only today got back the privilege of eating indoors (and only at 25% capacity). The governor wouldn't have granted that, had he not been sued (twice). The case count has been flat since June:
There's a really good FT article on this here: https://www.ft.com/content/b9653470-8779-4037-86ad-96edfcd6f...
Overall, Madrid opened much quicker and is now in the grip of another outbreak.
See https://www.cebm.net/covid-19/spains-outbreak-updated-on-reg... for some discussion (two weeks old though, so beware).
The third wave will probably be a lot easier to model (god, I wish I was joking on that).
Some people at my institution in fact have argued that you should not use flu-derived models, but instead model the spread and the outbreaks in similar ways to SARS and MERS, which rather than big "waves", had a series of spaced outbreaks. At an internal seminar they showed that at least their models for the Bergamo area seemed to be consistent with the actual data.
The U.S. is a very different picture, in many states the first wave never really went away.