Ranka recently reported that the World Amateur Go Championship used the MacMahon system for its pairing algorithm (
http://ranka.intergofed.org/?p=12658 ). This pairing algorithm was originally named after Lee McMahon (
https://en.wikipedia.org/wiki/Lee_E._McMahon ). From what I understand, he used this approach in a sort of club ladder setting in New York, whilst the British Go Association used it in a tournament setting, to replace Swiss. This is used by some to suggest that the name of this system should be McMahon. However the German player Christopher Gerlach, has a computer program called MacMahon. As this is the definitive version of the pairing algorithm, I believe that he suggests it is correct to now use MacMahon, even if he should perhaps have called his program McMahon in the first instance. This is because not all pairing algorithms are the same. (I am not entirely sure what the difference is between some FIDE standard of Swiss and McMahon, but I take it that there is some genuine important difference.)
A further complication is that McMahon, MacMahon, and indeed M'Mahon, are all effectively equivalent when you look back through time. Whether you want to call it the Ellis Island effect, or just say that it is was styled according to the fashion of the times, it really boils down to the same thing. Johnny may have chosen to write M'Mahon, but he could have equally written McMahon, or added an extra a like the Priest did. This of course mirrors the confusion in 1970s Europe over how M*Mahon should be written, with official BGA publications regularly using MacMahon.
Of course today, everyone LOVES standardisation, or is that standardization, so which should it be?