It's not exactly the same though.
It appears your friend has severe colour blindness where as mine is very mild... yet I can't get 100% on the numbers test. We had an afternoon at work where we did a few mock tests and I said a colour looked more orange where someone else said it was closer to red (although the colour could probably be seen as both). Another one is greens and browns where the shade of brown is so close to green that different people would name either. This isn't colour blindness, it's simply a matter of opinion on what that shade is. Railway signals are clearly a colour, there's no chance of mixing them up with oranges or browns. The rules for the test have been as they always have because nobody has ever thought of changing them to actually be better.
The biggest issue I have with this is I can stand on the far end of the platform and tell anyone what colour the signal at the other end is. I could do it over a million times. I can
prove I can tell the right colour, yet "they" can't prove that I can't.
The Ishihara test is flawed for those people like me who have a form of mild deuteranopia. The
Wikipedia article on the topic makes interesting reading.
I should think though it all depends on the medical examiner on the day. If you get a nice person, so long as you can reach the pass mark (even if you take extra time), they'll be happy.