Earlier quoted context omitted.
I think 2-2.5% frequently cited is the estimated IFR in the US.
I've only seen 2-2.5% cited as a CFR (case fatality rate) for 1918 influenza in the US, not the IFR (infection fatality rate). I don't know if those are assumed to be equivalent for are assumed to be retrospective analysis, but the CFR rate and IFR remain significantly different for COVID-19- ~4% for the all up US CFR rate and estimates anywhere from 0.1% to 1% for the COVID-19 IFR in the US.
That may be the case; the places I've seen it cited it's clearly a US number, and it's usually alongside global infections and deaths, but the denominator of the US number has generally not been clear.