It's not that simple. It's not even just about giving plants energy (although that's the main goal).
The key points to make are:
A) Photosynthesis isn't the only way plants use light
B) Plants can actually use the supposed "unabsorbed" wavelengths for photosynthesis quite well.
The figure to look up isn't chlorophyll absorption spectra, but photosynthetically active radiation (PAR):
So it becomes clear immediately why red dominates grow lights: it's very efficient, from a photon energy perspective, to drive photosynthesis with red photons. There are a couple problems with driving it with pure red light, however. One, in pure red, auxin synthesis goes crazy. Good for when a plant is ready to flower, but otherwise it causes excessive elongation and leaves your plant unhealthy. Two, red LEDs aren't nearly as efficient as blue LEDs. While the higher energy efficiency of driving photosynthesis with red wins out in that regard, the need for auxin inhibition means that you need significant blue in a fixture.
As you can see from PAR, colors between red and blue are
not actually bad for driving photosynthesis. In fact, if you add some green to intense red-blue light, the green drives photosynthesis *better* than the red and blue per unit energy***:
https://watermark.silverchair.com/pcp034.pdf?token=AQECAHi208BE49Ooan9kkhW_Ercy7Dm3ZL_9Cf3qfKAc485ysgAAAaAwggGcBgkqhkiG9w0BBwagggGNMIIBiQIBADCCAYIGCSqGSIb3DQEHATAeBglghkgBZQMEAS4wEQQMfFDsYGfmSAqMhXtAAgEQgIIBU4Exn7hx-gf2Acvp3bVw5i_UETVBu6u-5fkYoruDvNUeGUJtWhiIXJoMEFbEGmFXE2ZBEkqF-TVIngrEXZ3QsOpiZVx9VZ5cK0sk6xCN4Xgrf7PLAQ6tpnnsQv06k7jN5bhDo4l1UemnlaL1UhEePB1y6rorCV-VOP2uHfzgmATtGxdSt_qjjgskldaQmvBll5pI-8Z1FWGWwNFMVSvmUAz9YIQOcT9YSXopZ8NOe7lT5xhNscyKceJBlSDyF5DJ_0T8m_h4SeAQ9v0yzfN310a90fODB-y7nAFw9gqZ59eC5-6Q0wKqW0PCvIq8-68DOA8BN58JQ2sJZhTF-Er8Z3bfpIJa6zvksvujocceWhd99WveOiFKmO2K7AcThBePueRB8xF466QPRsoWVvUdFNouEr_kEFte4tFw11CdONoWbh0LIKfZJ6y9rn04tiVJ9OrsKAIn red-blue growth, the blue photons penetrate well (but are inefficient), while the red photons are efficient (but penetrate poorly). Photosynthesis rates on the surface become saturated. When green is added, it is more efficient than the blue, while it penetrates to a deeper depth (less saturated) than the red - to a depth where photosynthesis isn't saturated. Green's weakness of being highly reflected by plant leaves is reduced in nature by the interreflection between leaves when you plot it out on a whole sky-sphere perspective. People have for ages been growing plants under HID light, which is rich in green, yellow and amber photons. It demonstrably works - and this accounting for the fact that the means of generating photons with HID isn't nearly as efficient as the means in LEDs.
Green, yellow and amber all additionally have hormonal influences. For example, one of the reasons that you don't grow plants under LPS (only HPS) is that the narrowband yellow light overtriggers a response of storing starch in leaves rather than having it transport down to the roots. There are a number of spectral responses developed by plants for various reasons (the most important, however, being the red/blue auxin response, which is to respond to changing seasons - there's more blue light in the summer, more red in the fall).
The reason you don't see much green, yellow and amber in grow lights is because LEDs in these ranges are horribly inefficient. To the point that some white LEDs (aka, blues with a glowing phosphor coating) produce more green photons per watt than green LEDs do! Hopefully this will change at somepoint.
When it comes to red, while deep red (~660nm) would be more efficient at photosynthesis, the shorter wavelengths are more efficient at creating photons, so they generally dominate.
Some people will argue that NIR and UV are worthless - UV because it has to get way downconverted for photosynthesis, and NIR because it can't be used at all. I do not find this to be supported by the evidence - but it depends on the plant species. A fun experiment (which I've done) is to buy seeds of a colorful lettuce variety, grow some completely without UV, and the others with significant UV. The ones with UV will be brightly coloured. The others will be green. The colourful pigments in lettuce are basically sunscreen, produced in response to UV exposure. Plants have a number of such "sunscreen" compounds, and don't produce them when not exposed to UV (which is one of the reasons why you have to slowly "harden" plants to sunlight after growing them inside). Now, one could argue that such compounds have no utility indoors, but in addition to being visually appealing and promising in terms of health benefits, it doesn't take much energy to create them. All UV combined in sunlight is only 10% of the energy. Adding a couple percent of your energy as UVA seems worth the cost to me; UV LEDs are pretty efficient. As for NIR, some studies have indicated various plant hormonal responses to it of significance. It's probably not worth the energy cost, mind you, but it's not "worthless".
In summary - your best results are with:
* Lots of red for efficient photosynthesis driving
* A good bit of blue to prevent excessive elongation
* Some whites or amber-to-green LEDs to broaden the spectrum for hormonal reasons
* A couple percent UV LEDs to trigger the plants' "sunscreen" response.
*** Caveat to the green study: if you're undersaturating your plants (not giving nearly as much light as the sun), red will probably continue to be more photon-energy efficient than green (ignoring the efficiency of LEDs used to make the light). Likewise, if you gave an intense amount of green, you would saturate photosynthesis in the middle layers. Caveat emptor
But it's a moot point because you can't buy efficient green LEDs.