Yes, of course it would be nice if all inter-banking systems would support UTF8 or at least ASCII. But they don't. There are various inter-banking systems that simply are so old that they only support EBCDIC. Newer do, but even quite modern SWIFT still requires the EBCDIC set to make sure data goes through to each and every bank on this planet.
And in practice, if you have diacritics in your name, and do any kind of international business or travel, you are used to having found a transcription to use.
My wife has a diacritic in her last name. Even if forms at airlines now support entering them, she NEVER does. Because you can be sure that sooner or later in your travel, with this name getting copied to some immigration system, Covid health certificate database etc, it will get corrupted.
She never had any problems whatsoever with having her name transcribed on all travel documents, but with her passport containing her name with diacritics. Because, surprise, while I don't know what exact encoding the digital part of the passport contains, and least the machine readable line also uses only the ASCII or EBCDIC character set. So the best way to handle this technological debt is not to sue someone, but just use whatever that machine readable line in your passport days.
Names suck, anyway. The idea of having NON-UNIQUE identifiers to uniquely identify human instances is stupid anyway. Everyone on this planet should simply get a UNIQUE name in an universally agreed character set. I'd be even OK with Base64 of UTF8 ;)