My understanding is that the C++ source is required to understand Unicode (ISO 10646). For example 2.11 [lex.name] paragraph 1 reads:
An identifier is an arbitrarily long sequence of letters and digits. Each universal-character-name in an identifier shall designate a character whose encoding in ISO 10646 falls into one of the ranges specified in E.1. The initial element shall not be a universal-character-name designating a character whose encoding falls into one of the ranges specified in E.2. Upper- and lower-case letters are different. All characters are significant.
Whether it is an award-winning idea to create source code which can't be read by most programmers is a different idea. In general, security by obfuscation doesn't work: even if all your identifiers are essentially unreadable by many people, the structure of the program can still be decoded. However, I can promise you that I won't give any help with programs posted e.g. over here which I can't read: I'm already put off to some extend by identifiers using the ASCII subset of Unicode but a language I don't understand. If the characters don't make any sense to me either, I have probably no inclination to spend any time on the source. Sure, in the long distant past I had created a version of gcc which used German keywords. It didn't work too well because it already didn't understand the standard headers so I dropped it even before having written the obligatory "hello, world"-program.