I think rewording the CM spec to use the term “character references” is good, it would be more in line with HTML, and simplify phrases such as “Entity and character references”, or naming escapes and character references as two mechanisms similar mechanisms.
The HTML spec (no longer?) does not reference “entities”. Instead, there are two/three types of character references: named, numerical (hexadecimal or decimal). (for reference: https://html.spec.whatwg.org/multipage/parsing.html#character-reference-state)
I think rewording the CM spec to use the term “character references” is good, it would be more in line with HTML, and simplify phrases such as “Entity and character references”, or naming escapes and character references as two mechanisms similar mechanisms.