There's a few things that could be happening to explain the current high infection rate/low death rate we're seeing at the moment. It could be:
1. Covid has mutated to a less deadly form - this has been suggested as a possible end point since C-19 first appeared. If it is this, great.
2. Young people, having noted the low personal risk to themselves, are doing stuff that older people aren't and are spreading it around people who are low risk. If it is this, it's not good as it keeps the virus in circulation and could kick off a higher death rate later on when higher risk groups catch it from them.
3. Covid is a seasonal virus like other Coronaviruses - this was also suggested from the beginning and we might just be fooling ourselves into a false sense of security over the summer lull before it kicks off again in the autumn. If it is this, things could get really bad by the end of the year.
4. Something else
It might be one of these or some of these or all of these. We don't know for sure yet. We do need to be careful of putting our faith in one study supporting our favourite view. It takes numerous studies replicating results to get any sense of certainty. At the moment every researcher in a relevant area wants to be the first to have a game changing result and in the rush to publish mistakes can be made. Best to wait until subsequent studies support any findings.
At the moment I'm keeping an open mind, hoping for 1. and fearing 3. Only time will tell us what is actually going on.
I think you missed one of the more obvious options:
Due to lockdowns etc the number of people infected fell to a very low level.
With restrictions eased, levels of infection have started to increase in various countries.
Cases, especially mild or asymptomatic cases, are far more likely to actually be counted now because testing capacity etc changed a lot compared to the first wave.
So, even though cases are rising, they are still a very small number of cases at any one time compared to the actual number of people that were infected at any one time during the peak of the first wave.
And this low number of current infections is reflected in hospitalisation and death stats.
If numbers infected, and demographics involved, go back up beyond a certain level, rises in hospitalisations and deaths will be expected to rise too.
But due to changes in testing etc, we would not expect to see the ratio of positive cases:deaths to be the same as the first time around.
So if in March there were 5000 cases detected on a particular day and 500 deaths that day, and then in September we see a day with 5000 cases detected but only 15 deaths, I do not need to look far for explanations (these are made up numbers for illustration only). Because I do not make the mistake of thinking that because these were both days with 5000 detected cases, that the wave of infections was actually at a similar level on both those occasions. So I wouldnt expect the death rates to resemble each other either.
Anyway thats the sort of scenario I use as a foundation, and then various other possibilities can be tacked on top if some evidence emerges to suggest that there is more going on.
There are also moments where other factors can kick in which act as death amplifiers. During peaks in hospitalisations, even if system breaking point is not reached, there is a strain which can result in less admissions, worse treatment and more deaths. We are currently at the opposite end of things to that, far away from such a threshold.