It seems the strategy is/was to include any death as a Covid death if the person had previously tested positive within 28 days (Also some references to tests within 60 days) was done to ensure that deaths were not understated. I guess the flip side to that is that it then runs the risk of overstating things and therefore you then maybe default to excess deaths for a clearer picture. When trying to compare the data across different countries then you would need to know how those countries measure their own death rates, I’ve no idea how countries outside of the U.K. calculate their rates, do they err on the side of caution and count in similar way to U.K. or do they have a different measure which could see them count fewer (or more) than us. It does make comparisons very difficult.