I have the map I think you're talking about, and I've recognized various elements of it there, but Unkown00 obviously isn't trying to simply adapt that map to use NCS conventions; if they were, there wouldn't be any distinction between, say, Irish Americans and Anglo-Americans, because they both speak English.i think it's based on one of the older worlda language maps from the wiki
Still some things are wierd though. I'm not from the US but I'm sure Finnish-Americans or German-Americans by large do not longer speak their original languages or identify themselves by their original nationalities. And I'm absolutely sure Brazilians consider themselves different from Portuguese and Mexicans from Spaniards. The way every nation or country or culture distinguishes ethnicity is also complex. Self-identification and language usually is a safe bet, but how there are places that overall speak the same language but are multiracial and/or multireligious, other cultures that are similar yet different or are in different countries (Belgium for example, are the Francophones there Walloons, French, or Belgians, maybe all three?) and so on.So for sources I just Google random ethnic maps and I look at a few of them for the general idea. As for the language map, I use it either as a base map or I assume they overlap with ethnicity. Some like the Arabs state I used as a based map as I am aware that not all of the area is inhabitants.
oh and the borders of the italian states on the 1815 to 1885 maps are inconsistent with the rest of the mapsTwo things I have noticed and are mildly infuriating are the fact that Moldavia's northern border and Courland's borders in the late Medieval/Early Modern maps are wrong. Why are they different than the post-1900 maps?
None of the states that had seceded (save for West Virginia, which of course counter-seceded from Virginia) had been allowed representation in Congress at that point, so they should be colored as territories.