Space is always a factor. Also when you do have unlimited. With a limited budget and unlimited space the choice is really easy: get the most amount of light you can get for that $10k so you can use as much as possible of that unlimited space you got. Electricity is cheaper than the bud it creates. With unlimited space but a fixed budget for light, the cheapest light wins.
In reality you don't have unlimited space and have a given limited space for which you choose lighting.
Also, for such a grow op it would be an easy choice between cheap hps diy and 1k DE fixtures and bulbs. So it would be more realistic to compare with modern hps and not like the led fanboys here have done for years with a cheapo crappy philips knockoffs.
The white cobs are particularly useful for running soft and essentially allow paying for a higher gpw. The more you invest in cobs for a given space the lower the wattage needed for a desired amount of light, and the higher the gpw. No idea why that needs so many threads... the rest is fluff. People even want sub forums in the sub forum now, because grow journals with leds are special... If led is so great, why the fanboyness.
Are they worth it? Depends on what worth it means. They are above all worth it if you want to increase gpw by lowering the w. Which doesn't have to be all about money. Just as space is a limit, for many that goes for electricity use too. Some people can pay a lot to cut electricity use in half, others end up using the same wattage but add an extra tent or closet and still remain under the radar.
Energy use reduction is a given, inherent to running leds soft. Rather than discussing it over and over it would be a lot more fruitful to have a thread with calculations and examples. Heck, one simple rough formula is all it takes. The kW rate and lum/watt efficiency would be the main variables. Add a factor to guestimate reflector losses on the hps side.
With a 1400ma, 36v, cxb3590, 56% efficient it roughly comes down to paying nearly double for the setup (compared to DE) and saving 33% in electricity. Obviously doesn't include temps and spectrum and is rough bibled ledlogic. I don't know what you guys pay for electricity but I would earn that back in two years and don't sell a gram of bud. It would however take a lot longer to save enough in electricity for me to start caring. Imo, the larger the grow op, the more sense it starts to make to become concerned with max efficiency / energy savings. Only then do the savings/efficiency percentages actually reflect values that become interesting in terms of money.
The setup costs should be reasonable and if there's anything you can learn from this forum is that reasonable is highly subjective.
On a side note, if you're doing 3 runs per year in a large op, instead of 5-6, then light should not be your main concern.