Over in the AskUbuntu meta, I found it worrying that many search engines direct people to duplicates because their titles are more searchable. Often (or so I guess from looking at the viewcounts), people visiting a duplicate page do not make to the leap to following the Possible duplicate link to the original, answered question.
Could duplicates just embed the actual question in the closed link to avoid visitor confusion? Like so:
In the above page, Upgrade Ubuntu 10.04 to 11.04 would remain the <title> and <h1>, while How to Upgrade from a very old release to the latest would be an <h2> (or perhaps just a <p>) visually styled like the title. The answers to the original question would also be listed below.
The core idea is that a duplicate question link is basically equivalent to a redirect, but retains the duplicate's question title for searchability.
I think this would make SE's duplicate system more user friendly, and better optimized for search indexing.
To address the concern that this would be confusing: It is my position that current system is confusing. In cases duplicate links are more visited than the actual answered question they link to, doesn't that indicate our search visitors aren't getting it? And are leaving the site before figuring it out? This would ensure that the first page visitors see when hopping off Google contains the best information for their query.
Obviously the current operation of the duplicate system is perfectly clear for an SE guru. But maybe not to someone briskly hopping through Google results.
