@ehecatzin
@Boris Gudenuf
Yeah, you both have managed to actually convince me that - for example - at least for a few cultures it'd be really nice if there was an option to live as a nomad until like classical or medieval era and you'd need to settle down only to advance to the renaissance. As I won't believe you can develop scientific method, gunpowder manufacturing, printing press, caravels and industry as a nomad

But for sake of Goths, Franks, Huns, Scythians, Manchu, Mongols, Turks, Magyars, Bulgarians etc it'd indeed be really cool if you could live just from the steppe for first few eras.
I'd hope it would be closely tied to terrain system, so such civs would be possible only on equivalents of steppes and savannah, with plenty of grass for horse grazing and natural development of the best cavalry in the world (not to mention - conditions naturally preventing rise of cities and promoting nomadism).
Just want to point out that your skepticism about the idea forced me to refine my thinking about it - as usual, it is the synthesis that produces progress!
A case could be made that
Humankind's 'Nomadic Start' in the 'pre-Era' they are calling the Neolithic could be extended later in the game in very specific instances with a wider Terrain Range.
For example, technically, all the 'Indo-European' cultural Factions were still nomadic or at least 'semi-nomadic' until long after 4000 BCE or the nominal start of the conventional 'Ancient Era'. Even later, large and important Cultural Factions like the various German tribes (Franks, Burgundians, Teutons) in the Classical Era showed 'nomadic' tendencies from a northern Forest environment, and, as mentioned earlier, the Gallic Celts in the early Classical spawned massive invasions/migrations into Roman/Greek areas, also from terrain/climate combinations not usually associated with the 'classic' Pastoral groups.
It also might be useful to look at the fact that after the Gallic migrations were defeated or turned back, they started forming cities: urban concentrations like Bibracte only seem to have started rising in the century or two before Caesar's conquests, so we have what might be a useful historical example of the 'conversion' of a semi-nomadic group into a 'Civilized' faction. Also useful for bringing 'nomads' into the game is that the Celts had already pioneered several important technologies without having Cities: the long iron sword, link mail armor, chariot and wheel technology superior to the Roman when they met (most of the Latin words that refer to chariots, wheels, carts, harness, etc. have been traced to Celtic/Gallic roots, so the technology and words were borrowed), soap, and were not only building cities but a pretty good road system, complete with exact surveys of distances between towns which were so good the Romans kept them after they 'Romanized' the province of Gaul.
Which might give us some Technological Clues: the pastoral 'nomad Civs/Factions' were very good at Technologies involving warfare (the composite bow, saddles, and original wheel/cart technology is also traced back to the Proto-Indo-European nomadic groups) and Trade - roads and distances.
As you pointed out, though, any Pastoral Faction has to be prepared to Settle Down sooner or later. Even the Mongols had to give it up when their opposition had efficient gunpowder weapons, because in addition to an iron smelter, you cannot put a gunpowder mill or gun foundry on a cart either!
In the
Humankind game we are (sort of) discussing, I'd argue that a variant Nomad Faction should be an option for the Ancient, Classical, and Medieval Eras. If you want to hang onto, say, the medieval Mongols through the Renaissance you could try it, but, like the Mongol Successor Great Horde in southern Russia, the 16th century is probably going to see you getting your horse-sitting Butts Kicked by Muscovite Streltsi musketmen that you really cannot fight effectively.