In the S/M/L chart (this one http://asmartbear.wpengine.netdna-cdn.com/wp-content/uploads...) you can see that at the right-hand side, the M (green) activation is pretty low and goes to zero where L (red) is still strong enough to be visible. But there's no frequency on that graph where M (green) is active all by itself. So you've probably never in your whole life experienced a pure "green" sense input!

CIE takes this into account when defining visible color spaces. Since you have 3 "primaries" you would expect the colorspace to be bounded by a triangle, each corner being a pure input of each primary. But you can see here https://en.wikipedia.org/wiki/International_Commission_on_Il... that the top green corner is missing. That's because there are no frequencies of light that would let you see any color that is "greener" or further toward where that corner would be.

OK one last thing before I get to my proposal :) Your eyes gradually adapt against the colors you are looking at. For example if you're wearing red/blue 3d glasses for a while, when you take them off things will look greenish out of the "red" eye and yellower out of the "blue" eye. You can play tricks like this http://wxerfm.com/blogs/post/jfrieders/2012/jan/25/picture-c... In fact it's possible to exhaust some of your photoreceptors, like the "spot" from a camera flash or the streak from a glare of sunlight.

So here's my idea: if you strobe a really bright red & blue light, enough to temporarily blind the S & L receptors in your eyes, you should be able to see a green that is more pure and intense than you have ever experienced, right off the CIE chart of visible colors! Who's up for experimenting? :)