If you are interested in plant yield above all else then monochromatic light will generally work well. Of course if your crop is something that needs to look green and compact, a cabbage for example, yield may not be all that you care about.

Experiments have been done. Yields under monochromatic light are as good or better than broad spectrum light sources when the lights are extremely efficient or are targetted for extremely efficient photosynthesis. This has been done with LEDs at various red wavelengths and with yellow LPS light. In one experiment, LPS lights produced the highest yield of wheat at any particular power level, simply because LPS lamps produce so much light (LPS lamps produce 40%-100% more light than any other light source widely available commercially), although that was in a lab and they might not have done so well exposed to pests and weather.

Etiolation varies between plants but can be severe with just a yellow or red monochromatic light, and the plants tend to be a weird colour also as they stop producing certain pigments which are not being used. With LEDs it is relatively easy to add enough blue LEDs to produce more normal shaped plants without losing yield. You can add supplemental blue light to LPS also but in practice HPS (not monochromatic but with nearly all the light in a relatively narrow part of the spectrum) turns out to produce better results on a wider range of crop plants, sometimes with supplemental blue light, sometimes not. Metal halides and fluorescents are really not in the running in this competition, the white light is an advantage to ornamental hobby growers but not to commercial growers. Metal halides have also been prized by cannabis growers for high intensity with the short wavelengths necessary for producing alkaloids.

With light at just a few wavelengths across the spectrum, such as a triphosphor fluorescent or a standard metal halide, nearly all plants can be grown successfully for cropping or just as ornamentals. Again, even though the plants are perfectly healthy and normal, they may still not appear in their proper colours under some of these lights. For example, if a particular shade of flower isn't lit by the wavelengths given off by the lamp it will appear dull and flat. If you are breeding orchids in your basement, this could be a serious problem, if you are overwintering a few palms it may be irrelevant.

If one studies the results obtained by scientists that have looked at this subject (see link below), then one will quickly see that in fact yellow and green light is actually better than some shades of blue and is actually more efficiently absorbed, contrary to what one would conclude if one were to only study the chlorophyll pigment molecule in isolation, which has strong peaks of absoption in the red and blue wavelength. Fortunately the plants have other pigments called carotenoids to gather up all of the other wavelengths, and pass along the high energy electrons to the appropriate receptors for the initial stages of carbohydrate synthesis. For that reason, it is actually the output of total numbers of photons that is the single most important limiting factor in ultimate yield, and the reason that conveniently available cheap lamps like a standard triphosphor work just as well or better than "specialty" lamps, and why yields are quite abundant under pure HPS which is quite high in yellow and orange. I am currently running Ushio 850's, but I have also grown with 4100K Halides, and Gavita HPS with excellent result. I used to run a very high intensity (9,000 fc.) garden, illuminated primarily with HPS, and supplemented with 4100K halides, supplemented with CO2, which was quite awesome to observe. It was about 135,000 lumens illuminating a 15 square foot garden all surrounded by mylar. Paul Mozarowski.

Here is a link that might be useful: Photosynthesis quantum yield, observe how efficiently yellow and green light is used:

My ongoing observations in my own garden, with K&B 3000K T12,
Phillips T8 830, and Ushio T8 850, suggests that 3000K is too warm, and chlorophyll synthesis is poor under 3000K. The addition of 1:1 830 and 850 T8 triphosphor lighting shows that 4100K T8 will work well, and 5000K T8 triphosphor will work well. At least in a rose cutting garden. Paul Mozarowski.