> A regular light bulb converts about 2% of its energy input into light, and 98% into heat. This isn't relevant to your argument, but that actually isn't true,

> A regular light bulb converts about 2% of its energy input into light, and 98% into heat. This isn't relevant to your argument, but that actually isn't true, that figure is about 6~8 percent. * The "luminous efficiency" as commonly reported is just the lumen/W luminous efficacy divided by the peak value of 683lm/W. The value of 683lm/W is the peak of human eye sensitivity curve and occurs at 555nm. The only light source that can achieve 100% "efficiency" by that measure is a 100%-efficient radiator that emits monochromatic 555nm (bright green) light, the best a perfect white light source can do is about 37%. (see [1]) * That "perfect white light source" would be a 100%-efficient radiator emulating the Sun spectrum bandlimited to visible light (zero infrared, uv and no other losses). * It's possible to make a light source that appears white to human eye and is more efficient than that by emitting mostly at wavelengths the human eye is more sensitive to, and emitting less at wavelengths it's less sensitive to. That's part of the reason why manufacturers often offer CFLs with high efficiency but poor color rendering and lower efficiency but higher color rendering. (And don't get me started on the color rendering index.) * That definition of "efficiency" is obviously not what most people think it is when they hear that word (useful work / total work spent), some prefer to call it "luminous coefficient". But it's not all lies, it was defined that way because it's easy to compute and understand and because this is the value of interest if e.g. you're making a LCD backlight or if you're designing an interior lighting setup. [1]: https://en.wikipedia.org/wiki/Luminous_efficacy https://en.wikipedia.org/wiki/Luminous_efficacy