Doing both things allows the maximum number of users to benefit. I'm pretty sure that's what you were getting at.

The diff convention of marking inserted and deleted lines by + and - provides a "channel" of info showing text comparisons. If the + lines were also blue-green and the - lines also red-violet then there's two channels conveying the same information.

Using colors or not, we'd definitely keep the +/- markers.

The highlighting methods are distinct, but line information is the same. The +/- notation will still be useful on a monochrome monitor or printed page. On a spiffy new 2500x1600 display, I'd personally prefer the color output to quickly find what's changed.

Certainly depends on the source. But I believe you are about right. I wrongly trusted memory for the stat. The figure I used applies to a subset, green perception impaired males, which is around 5-6%, depending on level of impairment considered.

As you note, including the red perception impaired adds ~2% of males. And about 0.5% of females are also color-vision impaired. Overall it is around 9% of the US population.

The prevalence of color-blindness varies by racial/ethnic criteria, and between groups male/female ratio varies quite a bit. Color vision is a very fascinating subject.