conref.dita editorial review

I have committed the overview topic titled "Use by 
reference" to SVN as a revision of conref.dita. 
 
Four items need particular attention. They are 
indicated in XML comments.
 
1. The question after the following 
paragraph:
 
      If the referenced 
element has a conref attribute specified, the
      
above rules should be applied recursively with the resolved 
element
      from one  referencing/referenced 
combination becoming  one of the
      
two  elements participating in  the next 
referencing/referenced
      combination. The result 
should preserve without generalization all
      
elements that are valid in the originating context, even if 
they
      are not valid in an intermediate context. 
For example, if topicA
      and topicC allow 
highlighting, and topicB does not, then a 
content
      reference chain of 
topicA>topicB->topicC should preserve 
any
      highlighting elements in the referenced 
content. The result is the
      same as if the 
conref pairs are resolved recursively starting 
from
      the referencing element. 

      <!--What do you mean 
"as if"? There's been no mention of the
      
converse. Should this say "the result is the same whether 
the
      recursive resolution of conref pairs 
starts from the referencing
      element or from 
the referenced element"?-->
 
2. The question embedded within the 
following paragraph:

      A key is bound to the resource addressed 
by the topicref or keyref
      in which it is 
defined, if it is provided. 
      <!--What happens if  none is 
provided? It can't be resolved? -->
      The 
resource to which a key is bound may be a DITA map or topic, or 
it
      may be a non-DITA resource such as a 
graphic or an object specified
      by an external 
URI.
 
3. After the last of the related-links items, a note 
that I was unable to make a link that would take the reader appropriately to the 
submap because of my uncertainty about the organization of the 
content:
 
      
<linktext
      
conref="../common/complexattributedefinitions.dita">Complex
      
attribute definitions</linktext>
      <!--Need to make link to this submap 
rather than to this stub topic.-->
 
4. Should there be changes to related-links reflecting 
Eliot's reorganization of linking and 
addressing?