When choosing between food and production, I typically pick whatever I currently have less of. As China, I have a lot of food, and the larger your population becomes the less valuable food is (because it takes more to grow each stage). Once your population becomes huge, how much does working that farm really get you? Its 1 hammer and 2 gold, but specialist eventually become worth 9 of key yields like culture. As China once you have golden age and empress love, can your growth even be stopped?
I can understand this position, but I think it's misguided.
Every yield becomes less valuable as the game progresses due to inflation. Four hammer tiles can be pretty strong early game, but it becomes pretty minor when buildings start to cost hundreds of production. Science on forests can be really strong initially, but really stops mattering once tech costs balloon into the thousands. So from that, I think food shouldn't be inherently more devalued than say production or gold, and China's percent bonus is just as good throughout the game.
Really, it doesn't make sense to compare farms to specialists. You make farms
for specialists, and no other base tile improvement does this better. That's not even a debate. Sure, you could stop working a farm to slot a specialist slightly earlier, but if you're growing as fast as China can then there's not much point. What you should actually be comparing farms to is other tile improvements like mines and pastures.
For those who don't know, every citizen working a tile consumes two food (specialists consume 2-8 depending on era, disregarding policies.). That means if you're working a tile with 2 food and 4 hammers, you're not actually getting any growth from that tile, and it won't support specialists. That means that every tile
not a farm or sea resource with building support is supremely bad for supporting specialists.
I need to wrap this up, so sorry if this sounds terse from this point. Just in a rush.
If you want specialists, you need to work farms early. That's a given. Cathedrals amplify something you need to be doing anyway, rather than make a bad tile good. So you hit the point where you've filled every specialist you can and you're still growing like crazy, so you work extra tiles. You could start working a mine, and in the short term that will be better, but as turns go by it always gets to a point where it would have been better to work the farm instead as this would have lead to more citizens, which could work both tiles. The thing about China is that your growth is so insane that you very rapidly make up for any lost production or gold, so when working a production tile, you need to ask yourself how it's contributing to your victory. Sometimes it's justifiable, as you're working on a specific wonder or building that helps you significantly, or you just straight up need GPT or you'll bankrupt. But unless you're near the end game, the long view says that growth will pay out better in the end, and winning a science victory is a marathon, not a sprint.
Couldn't figure out a way to work this into another paragraph, but it's also important to remember that, unlike every other yield in the game, the majority of your base food yields comes from tiles worked. Production, culture, science, faith, gold, all of these eventually get the majority of their base yields from specialists and buildings. Food is the sole exception to this as there is no Food Specialist.
Oh and to answer your question, I will sometimes just set cities to growth, but I typically micromanage them until they grow every 2-3 turns.