Update PR as suggested by Tim

Author: tomaszkam

xml/issue4469.xml

@@ -18,9 +18,8 @@ static_assert( identifier_of(p) == "os" ); // guaranteed?
 </pre>
 
 <p>We should clarify that parameter names used by standard library implementation
-remain unspecified, by making behavior of `has_identifier` and `identifier_of`
-unspecified on reflection of such parameter. Marking the behavior
-rather than result unspecified, allows implementations to reject such calls.</p>
+remain unspecified, by making result of `has_identifier` and `identifier_of`
+unspecified on reflection of such parameter.</p>
 </discussion>
 
 <resolution>
@@ -36,8 +35,10 @@ This wording is relative to <paper num="N5014"/>.
 Unless <i>F</i> is designated an addressable function, it is unspecified if or
 how a reflection value designating the associated entity can be formed.
 <ins>For any value <tt><i>p</i></tt> of type `meta::info` that represents a
-reflection of a parameter of <i>F</i>, the behavior of <tt>has_identifier(<i>p</i>)</tt>
-and <tt>has_identifier(<i>p</i>)</tt> is unspecified.</ins></p>
+reflection of a parameter of <i>F</i>, it is unspecified if <tt>has_identifier(<i>p</i>)</tt>
+returns `true` or `false`, and if <tt>meta::has_identifier(<i>p</i>)</tt> is `true`,
+then the `NTMBS` produced by <tt>meta::identifier_of(<i>p</i>)</tt> and <tt>meta::u8identifier_of(<i>p</i>)</tt>
+is unspecified.</ins></p>
 </blockquote>
 </li>
 </ol>