So I never renamed this type to TOKEN! because I've been a chicken. 
Part of what makes me hesitant is the idea of people working on codebases where they are considering WORD! to be a token in their dialect, and so they would say token: 'foo.
Calling it UTF-8! might seem good, but all strings are in the category ANY-UTF8? (Should that be "ANY-UTF-8?", with the hyphen?)
And ANY-UTF8? as a category does not imply immutability. But these are immutable.
So...CHARS! or CHARACTERS! or CODEPOINTS!?
The term "RUNE" is getting some traction for meaning "codepoint" in Go and C# (excluding some illegal ranges like surrogate pairs).
We Could Call It hashtag!... But...
When we say:
>> second "abc"
== #b
Do we really want to say that "single-codepoint hashtags are our currency for characters"? 
I don't think so.
Maybe Just Go With My Gut... Like With "Keyword"
TOKEN! is what has been on my mind since the beginning of the unification of ISSUE!+CHAR!, and it hasn't gone away.
What we name our fundamental types isn't the end-all-be-all of how a word needs to be used.