The fatality rate is complex. Estimates take into account the fact that not all cases were reported. If you look just at reported cases, you'd get a fatality rate of 2%. Everyone agrees that this is too high. In the Spring the fatality estimate I believed was 0.5% or a bit higher. (Here's the most recent estimate I've seen.
https://medicalxpress.com/news/2020-10-covid-deaths-infection-fatality-ratio.html. It's about 1%.)
I haven't seen estimates for the current spike. My point is that looking at the reported cases and seeing that we don't see as many deaths as in the past doesn't work, because we're probably getting more reported cases. But that's looking at conclusions a layman draws from the reported numbers. The actual fatality rate that people quote isn't based on reported cases, but estimates of actual.
I'd guess that that real rate has also gone down somewhat. There are better treatments than in the Spring. But you can't use the case numbers you see in the press to verify that.
Hmmm.... here's something discussing the decrease over time:
https://www.snopes.com/news/2020/11/03/death-rates-have-fallen-for-hospitalized-covid-19-patients-as-treatments-improve/. However it may be slightly misleading. In the Spring hospitals were severely limited in capacity. It's reasonable to assume that a larger fraction of people who are infected are now going to the hospital. That would tend to lower the death rate in hospitals. That rate isn't quie comparable to the overall IFR, the fraction of all people infected who die, which is what most people mean by death rate.