For people with protanopia (absence of "red" cones), confusion lines look like this (which is quite intuitive for me)
Analogous picture for tritanopia (absence of "blue" cones)
But picture for deuteranopia is quite different. Confusion lines are intersecting at one point. But this is not a point nearby green color, it is opposite. Why?
I have a conjecture. Let me remind you about RGB. There is a three-dimensional space (coordinates are RGB coordinates). There is a cone in this space which corresponds to all colors which we can possibly see. Then we consider a section of this cone by a plane. And this is your RGB diagram.
I think that people have chosen a random plane. There is a line for excitation of green cones only. And this the plane intersected negative part of the line. If we chose this plane more carefully, this the picture for green confusion lines would be similar to red and blue.
A friend of mine says that that choice of this plane is not random. It is a plane of constant brightness. He even conjectured that "green" cone cells are activated not when they absorb green light, but when there is no green color at all.
Is this plane indeed random?
For more details look up here