I'm something of a lighting nerd. I like the whole topic of lighting; not only is it the first invention that heralded the beginning of the age of electricity in human history, but it is also the one invention that permits pretty much all of modern architecture, manufacturing, and other technologies.
Without artificial light, you can't have big buildings. You can't move daylight very far into the interior of a large building; ask the people who work in the Pentagon, or in some of the middle stores of the Mall of America. Shoot, ask me - my office is a plywood shack inside the basement of a brick and concrete structure; to get to daylight I have to go through two doors. There are lots of excellent technologies designed to move daylight from outside to inside, but those only work so well, and the farther you get from an exterior surface the more academic they become. So for everyone who ever got to work before sunup or had to keep working after sundown, artificial lighting is a huge part of what makes your workday possible.
The original artificial lighting option was fire. For further reference, I recommend Artificial Sunshine Maureen Dillon, and I wrote about that very good book a couple of years ago. And fire, the assorted variations of fueling, flame size and carriers were the entire lighting gamut for the human experience for thousands of years. Eventually Edison and many many others came along and whipped up some alternatives that didn't require open flames.
They still did their thing by getting hot. A filament energized by electricity to become so hot that it glowed in the visible spectrum, was the device originally invented and demonstrated by Humphry Davy in 1802, but never successfully commercialized until Edison came along and threw his considerable efforts at the problem, culminating in the first practical, mass-produced electric light bulbs in 1879.
The very best incandescent lamp efficiency I can find is about five percent efficient - which is to say, of the power going in, only five percent re-emerges as visible light. The rest is heat. And that's for specialized projection lamps, which run crazy hot and don't last long enough to be practical lighting lamps.
Fluorescent lights aren't very new, either. The last significant development of fluorescent lighting technology, that of coating the inside of a neon tube with a fluorescing material, came in 1926. Since then, most of what's been going on with fluorescent lighting has been a matter of incremental improvement, both in the materials of the lamps themselves and in the control systems that condition and control the power that runs the lamp.
The most efficient fluorescent lamp I can find mention of is a new, high-output T5 tube generating light at about 105 lumens per watt, which is about 15% efficiency.
So you know, compact fluorescent lamps aren't as good as that, though they are good.
The very best LED lamp delivers light at about the same efficiency.
We can do better: the very best low-pressure sodium vapor lamp produces light with about twice the efficiency. Granted, the light it provides is orange.
The day is coming when we're going to have a light source that is 50% efficient - only half the energy that goes into it will be wasted. Look at it from another perspective: next time you're at the grocery store, look up at one of those fluorescent fixtures that has four tubes in it. Now imagine the exact same amount of light as all four of those tubes provided by only one of those tubes, and that tube still getting the same amount of electricity into it as now.
That's going to be one seriously bright tube.
Just a couple of weeks ago, Cree announced achievement of over 270 lumens per watt in an LED lamp. In a real-world application you can expect that to be considerably less, but even so consider: to get the same kind of light as a fluorescent tube using 25 watts, I would need an LED tube using only 10.
That's pretty good.
I have a few more comments regarding lighting but that's plenty for now.