IIRC, that study processes the data to look at the percentage distribution of causes of deaths among total deaths, so it's not looking at how many people died, but rather if the proportions seem off.
To reach its conclusions, there were a some assumptions made about how things are supposed to look, so while it's intriguing, there are many way a skeptic could explain away the conclusions.
Hopkins retracted this: https://web.archive.org/web/20201126163323/https://www.jhunewsletter.com/article/2020/11/a-closer-look-at-u-s-deaths-due-to-covid-19
IIRC, that study processes the data to look at the percentage distribution of causes of deaths among total deaths, so it's not looking at how many people died, but rather if the proportions seem off.
To reach its conclusions, there were a some assumptions made about how things are supposed to look, so while it's intriguing, there are many way a skeptic could explain away the conclusions.