No reason besides the fact that Unicode characters can cause countless errors in legacy systems downstream.
Perhaps the money that could be spent on supporting arbitrary renditions of indigenous languages in non-standard latin-derived characters would be more wisely spent on supporting those communities directly.
You can't honestly expect all of the people involved to learn to input all the possible chractersets that some groups use.
You could train them, but I think spending fraction of money would do massively more good in other ways...
All systems worthwhile have been using UTF-8 since the 1990s. There is no excuse for legacy systems handling business applications not to handle UTF-8.
That's irrelevant. You can rest assured that in any larger government system some information must travel through the equivalent of a chimpanzee trained in Baudot code.
It should have been built to support UTF-8 in the first place and should be upgraded, but the time it will take to do that will not help this person. With any luck, it may help someone else in the future though.
My point is that there are plenty of reasons why this kind of name is not supported. Besides the legacy computing reasons, having a name that only a handful of people in the whole world could spell out is going to cause problems.
Fixing this problem would cost real money that could be used on other things. Perhaps the indigenous community should be consulted on their most pressing issues before enshrining the right to have these characters in a business name.