On the surface it would seem that with Democrats winning such high percentages, there is no way the GOP would ever win the popular vote. But the substantial decline in turnout in the South's general elections meant that the mid 50's numbers the GOP would get in the big Northern states would be sufficient to win the popular vote as a result.
This is why it makes no sense to calculate nationwide popular vote in the US - or, at least, to take it is a measure of nationwide popular will. States have different electoral laws. The president is not elected by the popular vote. If s/he were, turnout patterns would have been very different.
In any case, adding popular vote in SC in 1924 to the popular vote in New York that same year is akin to calculating an average body temperature in a hospital (including the morgue): pretty damn stupid exercise.