There really should be a consistent criteria for civ selection, but it seems like no standard is ever followed. Ideally they should follow some guidelines like this:
- Exist for the majority of the game’s timeframe (~1500 to <1900)
There needs to be some bounds on what countries should be included. Most people would be upset if they included civs like Byzantium or the Soviet Union.
A number of civs violate this with Aztecs, Mexico, and USA being the worst offenders (Inca also doesn’t in its current depiction but if they did persist as an insurgency long enough to potentially fit this criteria).
- Have a suitable military that can fill a unit roster
This is where having a coherent timeline impacts gameplay. There is an established progression from archaic units to modern ones. Civs like Mexico and USA are too modern to have archaic units, and civs like Aztecs lack anything modern which makes balancing a struggle.
Malta lacks this currently, but a lot of that is because they were very poorly designed.
Other civs like Lakota stretch this requirement with their total lack of artillery.
- Be a distinct and independent state or cultural group of some significance
USA and Mexico violate this because for half of the timeframe they are indistinguishable from colonial Britain and Spain. Implementing them as a much more developed revolution mechanic would have been far superior to a standalone civ.
Malta violates this because they were never really independent and don’t represent a significant cultural group.
India and Germany also violate this by being nebulous umbrella civs. These could be fixed by breaking up the civs like AoE2 did with “Dynasties of India”.
Some “insignificant” civs like Hausa and Lakota are good representatives of their cultural groups and had periods of prominence like the Sokoto Caliphate and Battle of Little Bighorn.