I, for one, would really appreciate it if people would take a little more time when considering duplicates to name or link to the node that is being duplicated. I see several advantages to doing this:
Each voter can verify that the node is in fact a duplicate according to her or his definitions of duplication. For example, one of my nodes was considered as a duplicate because someone else posted an answer while I was composing mine. It's also possible for the word duplicate to mean 'someone asked this question just yesterday.' Neither of these are what I consider duplicates (to me a duplicate is when the same person posts the same thing), and I would vote to keep both of these nodes.
In verifying that the node is a duplicate, each voter can also vote responsibly. Right now it would probably be possible for me to consider a perfectly fine node, say it is a duplicate, and have it deleted (although it would be reinstated later). It wouldn't have to be malice, it could be a simple accident. There are also people who, when they feel they have insufficient information, vote to keep or edit the node. In the case of a real duplicate, these votes could prevent reaping.
At times both the original node and the duplicate are considered as duplicates. It should be clear that they are considered in relation to each other--so that people know one consideration was accidental--rather than in relation to a third post which doesn't actually exist.