Here is me rambling for little bit.. Plants do appear green because a good amount of green light is bouncing off them. However again this explanation is grossly over simplified.
The truth is plants are pretty inefficient at absorbing/utilizing light energy in general. In best case scenarios only about 30% of photons are absorbed (Other inefficiencies within the plant means that even less is actually converted into biomass). This means a "green" plant under normal conditions is also reflecting back lots of purple, blue, yellow, orange, red light. However, our human eyes only have receptors for red, green, and blue. Since the chlorophyll does absorb a greater portion of the red and blue, the plant appears green... but this is a physical limitation of the human body not the plant. When the chlorophyll breaks down (think of tree leaves in the fall) we see the leaves change color because less of the red and blue is being absorbed by the leaf. Under absolutely perfect lighting conditions, the plant should appear black because it's absorbing all of the light.
LED is the first type of lighting that we can use to target specific wavelengths. Research in this area is progressing rapidly, but we still haven't fully dialed in this technology. By providing light in the appropriate quantities, we make the system much more efficient by eliminating the wasted photons. However, plants are complicated organisms and therefore it's hard to determine the exact effect a specific wavelength of light has on a plant. Just because a plant grows a certain way under 100% blue light doesn't necessarily mean blue light causes that type of growth. Instead, that type of growth might be because the plant is compensating for something it's lacking.
We know that all of the hormones in a plant work in at least a 2-way balance to regulate plant growth. For example, applying just auxins or cytokinins alone will produce very noticeable and specific changes in growth. So at first glance it's easy to attribute that growth to either auxin or cytokinin, but further research has revealed that it's actually the ratio of the two hormones that determines the outcome. Likewise, we have to be careful to avoid this in assessing how a specific wavelength effects the plant.
Another unique feature of LED that is just starting to get some attention is their ability to turn on and off rapidly and efficiently. If you've ever played with those remote control RGB lights, they run off of PWM (pulse-width modulation). It's basically a timer that allows the diode to switch on and off rapidly. The advantage is that you can blink a LED on and off so quickly that it appears as continuous light, just slightly dimmer.
In my opinion, this is where we will find the next big advance in grow light technology. People like to think of plants as machines... if we use more nutrients, water, and light it should grow bigger/faster! All of us learned early on that's simply not how things work. We know it's possible to photosaturate a plant, where any additional light is wasted. However, it's quite possible that we might find certain "flash patterns" that overcome these obstacles. A simple Arduino or RPI controller puts the technology in everybody's hands.
These are all purely hypothetical scenarios, but an idea of what could be possible in the future... For instance, maybe running a strong light, that switches on and off every 1/4 second might produce the same growth as continuous lighting by allowing the plant to "rest" in those 1/4 second off periods. The light would then be 50% more efficient! Another possibility would be to flash 2 different wavelengths independently. In this example, you could flash blue at one interval and red at another. There's a lot of room to experiment whether it's best to over lap them, or run one while the other is off. I find it completely ironic, that a flashing blue and red police light could potentially make a great grow light.
The last scenario is to try and mimic a normal daylight cycle. This would be a bit more complex, but you would have mostly red starting the day (sunrise) with more blue filling in through the day with almost all white (RGB) by noon. Then you would start to taper the blue off again so that it's mostly red/farred at sunset. If any of these lighting conditions produces good growth that is on par with "normal" growing conditions, it could greatly increase the efficiencies of our lighting systems. Within the next 10 years 2g/w will be the new standard.. not because of greater yield but because of wasting less photons.