The comment itself should be sufficient -- if you use the UTF-8 characters you're just introducing cross-character set headaches and wasting time looking them up that could be spent writing better code.
If you're using the conventional character based ones half the time someone's just gonna go "what are these junk characters" on anything more complex than the simple smilie and delete them.
But I'm old school, even if my content is going to be in a wider character encoding, I prefer to keep my code within the 7 bit ASCII limits just so I never once have to worry about my source getting boned by a character encoding mismatch; since pretty much every character set has ASCII codes 0x00 to 0x7F as its base.
But what do I know? I still say 99%+ of English language content has zero business using any characters outside ASCII-7! Oh noes, eyes cantz usededed teh styled quotes -- big deal!