The "universal" in unicode went from meaning "you can represent any modern language in it" to "you can translate any encoding in it without loss of information

The "universal" in unicode went from meaning "you can represent any modern language in it" to "you can translate any encoding in it without loss of information AND any character you could possibily want to have in a font will have it's codepoint". And so now you have precomposed accented characters and combining diacriticals, presentation forms for arabic characters, precomposed ligatures, double width latin characters (that look better with CJK characters), icons for chess pieces, emoticons and control characters (including several different spaces and newline characters). Actually supporting all of unicode is a massive undertaking, and if you start using lesser known features of unicode you will find plenty of bugs everywhere.