Sweden opted to take on more cases, and this may have lead to a greater percentage of infections in the most vulnerable populations, hence perhaps a high mortality. It is not clear (to me) that the Swedish health care system failed to hold up, although certainly if vulnerable populations were hard-hit that would indicate some policy failures.
It is also not entirely clear that the mortality is as high as claimed, as we do not really know what percentage of the Swedish population has been infected. I expect this will clarify in the next several weeks, as more antibody testing is performed.
As to whether the greater exposure leads to an improved herd immunity effect (as compared to other countries), we likely will not know until the end of 2020 (unless we learn sooner than that that immunity simply does not last very long, but this seems unlikely based on evidence thus far). Similarly, we will not have a clear indication of overall relative mortalities before that time.
There is the obvious counter-balance that Sweden's economy has almost certainly taken less of a hit than the economies of most other countries. But so far as the virus impact goes, I really would not draw any hard conclusions until more data is in. It might turn out to be a disasterous strategy, or it might turn out well, or it might be somewhere in between, as is often the case, when all options carry serious drawbacks.
