However, a fact is that a connected 2nd city gains a lot more raw yield than growing capital does (well, unless capital has say 4 excellent tiles, like double pigs + double gold). Recent NC Mao is a good example.
Almost any whippable capital (2 land food sources, maybe some early happiness near) does not need to fear double worker opening to be crippling. Yes, yield-wise there can be some better maths, but if we put value on the increase in flexibility and generally better efficiency with earlier workers, earlier roads, etc., it becomes very hard to evaluate how it translates into a win, but it is easy to see how extra safety due to ability to burst produce with chops can save the game from barbs.
I looked at the NC game, and due to very slow starting capital, I'd also go for settler before 2nd worker, especially because it is not many worker turns to connect it.
What I like in some general case. Faster capital with 6
tiles, 5
can work too will produce both the workers and settlers faster than NC Mao map. You set the capital for whipping with 2 workers, whip the settler and soon hit the happy cap but you don't care as your 2nd or 3rd city will share 6
tile and whip their own stuff. You start whipping and using that unhappiness timer sooner, which is also a resource of its own, especially if no early happiness is around. This is often how my starts looked like. Ok food, bad happiness (as I selectively declined gold/gems starts).
The main message, from my view of mediocre microer having played dominantly on partially handicapped Standard/Medium_sea_level/Fractal Deity, is that double worker is sometimes good on Standard/Medium_sea_level/Fractal Deity. So, the settings normalized answer is - depends.