You'd think so but not necessarily.

Photovoltaic cells can take in a broader range of energy from the sun than plants can through photosynthesis. LEDs can then emit that energy at specific wavelengths such as red or blue that the plants can take in.

The whole system can end up being more efficient.

Furthermore the prospect of growing plants in a completely controlled environment where you can control all the variables from light intensity, light cycle, temperature, humidity and remove the pest factor and therefore the cost of pesticides can result in a much better final product.

Honest question: Would this type of production be considered organic? There's no pesticides and such, but it's also not quite "natural".

I think it depends on which definition one subscribes to. Broadly speaking, there are two major lenses through which "organic" is viewed:

1) Focuses on holistic management, rejects intensive injection of industrial inputs into the farm/garden, and emphasizes applications of emergent activity of native species rather than aggressive "pest control".

2) Focuses on keeping "unnatural" chemicals or genes out of / off of the food.

I think this approach would tend to look organic under lens #2 but not lens #1. Various definitions of "organic" lend different weights to them and introduce other considerations, though.

I think that you're right. AFAICT artificial illumination for growing plants makes sense for growing staple foods in deep space if crewed missions ever get out there, or high-value specialty crops (delicate fruits/greens/herbs, cannabis) near terrestrial demand centers. It's worthwhile to grow tomatoes artificially in Edinburgh in the winter if you have customers who will pay the premium over bred-for-transport-durability tomatoes shipped in from abroad.

A really good diode light source might have a wall plug electricity-to-photons efficiency of ~50%, and the best rooftop solar modules achieve a bit over 20% light-to-electricity conversion efficiency. I would really be curious if anyone can provide citations to the effect that sunlight -> PV -> LED light -> plants is more efficient than sunlight -> plants. You'd have to get an enormous benefit from wavelength shifting, enough to offset the ~90% energy losses from sunlight -> PV -> LED light.

We have 46% efficient solar cells that are not currently cost effective. 45% of sunlight is in the useful wavelengths for plants, but the important comparison is their peak effecency with the optimum wavelength of 30% vs there theoretical limit with sunlight of 11% which is a ~3x boost under optimum conditions. Another consideration is leaf peak obsorbtion is around 10,000 lux though trees for example capture light through multiple layers. So, some plants are going to be more efferent at lower light levels. (this is also why trees don't look like a mushroom with one big leaf for maximum light capture.)

So, ignoring costs you can come surprisingly close to break even in terms of energy budget. Though we are likely better off going with GMO plants to capture more light than build millions of acres of solar farms just to break even.

With solar panels and LEDs you can shift the spectrum of incoming light into the ideal range for photosynthesis. My understanding is that this outweighs the conversion losses.

Many locations have too high a population to feed locally, and eating locally appeals to many people. LED farms in basements could help with that.