The mine is the problem, not a solution. Light is the limiting issue. Algae could work on the surface if a suitable highly productive strain could be found. If the algae were perfect except for susceptibility to contamination, then a surface system in the desert with greenhouses or glass tubes might work.

Suntubes with surface sunlight concentrated with mirrors would probably be even less efficient than electric lighting since you would loose about 10-20% per meter due to reflection losses on the walls of the tube.

Anything you can do in your mine, I can do for a tenth the cost in a corrugated steel silo.

Even so, land is not the limiting factor (in most countries) for a massive industrial project like this - it works in the middle of nowhere even better than it works in the city, and it is highly mechanized.

I don't have numbers on overall efficiency of LED lighting, only that the cost per lumen for white light is so high (in lumens per dollar) that LEDs are practically useless for bulk lighting. Compared to arc lighting or flourescents, they cost 50-100x as much, without even matching efficiency. LED evangelists refuse to compare energy efficient lighting, and assume that their trends will continue forever.

Currently, the best choices for bulk lighting are all high intensity discharge: metal halide, sodium vapor, and sulfur lamps. You can pick up a 1 kilowatt MH lamp for 20 cents a watt, and it will give you 60-90 lumens per watt of high color accuracy white light, and last you 10,000-20,000 hours. None of these come in small sizes, and they are all delicate systems including a ballast, igniter, and warmup/ cooldown time. And sodium lamps are yellow/orange.

High efficiency flourescents can match them in most things, and beat them roundly in modularity + ease of use.

LEDs are excellent for portable/fixed colored directional or colored flood lighting, good for portable white directional lighting, and horrible for anything else.