I encourage people to scroll around Google Maps (or better yet, drive) through some of America's most productive agricultural regions. The sheer scale of American farmland dwarfs any possible urban farm development.
I know from personal experience growing seedlings in my basement here that new LED grow lights make it much more cost and energy efficient to run a greenhouse than it used to be. But fertile soil is a massive input cost that you can't really get around for most full sized plants. I spent hundreds of dollars recently on garden soil for a small set of raised beds. I have a tiny flock of six chickens supplementing my compost pile.
Anyone can grow bad produce indoors. I think the real problem (having lived in urban, rural, and suburban areas) they're trying to address is the transportation cost of bringing quality produce to the city. Right now many cities are not really set up for affordable freight transportation. Cities like NYC are almost designed to make it pointlessly expensive and complicated to bring freight in from the country to the store. Commercial truck traffic gets redirected to cramped streets while personal car traffic gets access to the expressways and parkways. That's one of the reasons why it costs you 4x to buy a pound of worse quality bacon in NYC what it costs in many other US towns and cities.
So if you are trying to deploy technology to make quality food more available to American cities the solution is probably not just hanging up some grow lights and growing weak plants in trays. It is to attack the largest sources of the added costs. The 'local' label has salience to high end grocery shoppers because they believe that it indicates quality. If 'local' comes to be associated with tray farms using poor inputs, it will stop being associated with quality in the minds of Whole Foods shoppers.
It'd be great if America had the sort of urban freight network that was state of the art in Japan in 1980 instead of trying to come up with pretend-technology patches to much larger cost of living problems.