This is the bohemian play rate and win rate by patch. They had only 1.6% pick rate, despite being n1 in win rate, but then the playrate grew to 2.2%. The reason it got so high back then, is because of the Phosphoru rush popularity, and even then, it was “just” a 40% increase, but it was during the time we got new civs (Georgians and Armenians). And also, these screenshots go against most of your points - did you research them?
The moment Armenians were released, they had more than 2.2% pick rate, and it was 2% for Georgians but they had the lowest win rates of all civs. The moment Georgians got tweaked and sprung to n4 in win rate, their playrate increased by 50%, but it was still much lower increase compared to the initial drop after release, and did not even reach 1.5%. And the reason for the the initial pick rate was that they were new, and the playrates then drastically dropped.
The same would be expected of new civs - more or less the same play rate regardless of winrate. But I think that because most players who care about the nature of the game disagree with the 3k civs, their playrate lags behind expected numbers. I don’t expect Jurchens to have a pick rate close to Khitans, but I expect them higher than any of the 3k civs - i’d even say much higher. Let’s wait a few days to see.
Edit: also, there are 50 civs in total, and the 10 best performing ones have 21.94% pick rate in total. That is just 1.94% deviation of an expected pick rate in case of a perfectly equal pick rates among all civilisations - and this is carried by Khitans, HEAVILY. Shu are even below 2%. If your argument is that “the higher the win rate, the higher the pick rate”, then - apart from Khitans, it’s actually not the case, quite the opposite, since without Khitans, the bottom 9 of the top 10 would barely have a 17% pickrate - 1% lower than an expected value of a perfectly equal pick rate.