CEA (Controlled Environment Agriculture) is not bad. When over-engineered and quickly expanded without working on unit economics it might be a bad idea.
Traditional farming is not viable near the place of consumption. It needs a lot of land, and land parcels that size near a city is impossible to find. And even if you find, it would rather be used for more lucrative purposes such as commercial properties, than farming.
So to make farming viable near the place of consumption (there by reducing the distance produce has to travel, there by reducing cost of transportation and wastage during transportation, there by selecting seeds which are less hardy for transportation but more nutritious and tasty becomes a possibility), we need to improve the yield of the farm and the consistency and flexibility.
A. Yield of the farm depends on ->
1. space requirement between each plant (which depends on the ability for the plant to absorb nutrients and access to light),
2. amount of light (the bullets) and
3. the amount of carbon dioxide (targets). Photosynthesis is nothing but when the photons in the light break the carbon dioxide bonds and release the oxygen to the atmosphere and carbon combined with hydrogen from the water becomes hydrocarbons (the mass of the plant).
4. Quality (same size, no nutrient deficiency like tip burn or spotting)
5. No pest waste
6. cycles per year
In a hydroponics farm, since nutrients can be dissolved into the water uniformly the space requirement between plants is lower compared to traditional land based farming, the quality is uniform as the nutrients density in the water flowing is uniform, the light (including artificial lights can be increased), the carbon dioxide within the farm can be increased from 400 ppm to 1200 ppm (increasing the targets). More cycles in a year are possible, layers are possible.
With all the benefits, those farms near the city try to improve yield enough to make the 1-2 acre farm near the city viable (as thought it is a 20-30 acre farm 100s of km far away).
The savings is the transport cost, the wastage cost during the transport, the wastage during quality checking cost etc.
B. Consistency
Like mentioned about, good seed selection, uniform nutrient dosing and controlled environment so no pest attack means similar sized produce. This helps with inbounding for Retailers. They have to spend less time and money on quality checking or managing sell-able period.
C. Flexibility
In farming, big retailers have all the power. The contracts are one way forced. If they have a contract with you for 5 tonnes of cherry tomatoes and you aren't bale to deliver it on time, they will penalize you. But say you have 5 tonnes of spinach which you have harvested as per the contract, they can always ask you as a favor "Hey unfortunately our inventory is still not cleared, can you delay by 1 week". Now when you are running a farm at capacity, such delays are not easy to accommodate, because the next set of plants that need to be transplanted from the nursery are ready and you need these plants to be harvested out so they can be planted here. Harvesting and keeping it is also not an option due to low shelf life.
Here is where playing with light and carbon dioxide inside the CEA is super helpful. You can increase the CO2 and increase the light to speed up growth and you can decrease the CO2 and light to slow down the growth of the plants. And this flexibility means, you give more tolerance to the uncertainty in forecasting for the retailers. You take care of their headache. And this is valuable.
I run a pretty successful hydroponics farm in India and supply to online retailers and we are their preferred suppliers purely because we take care of their uncertainties. Ours is not super high tech. Labor is cheap in India. We have some essential tech, like the lights/CO2 etc. But that's about it. We didn't over-engineer.