As a user working in our JIRA instances for several years,
When I do a search,
I do not want to be inundated with hundreds or a thousand records meeting my search.

In this case, we have very few technical, power users; we have hundreds of non technical users who do searches to try and locate what they need. Therefore, they are not going to remember how to use various functions to locate the information. We want to be able to not allow our search to search across issues in archived versions so that they are searching more recent information. We do not want to move issues from one project to another archived project because:
1) You have to setup all the versions, components, issue types, screens, in the archived project to be able to support the move of thousands of issues from one project to another
2) You would have to create another field to populate (which, if closed, you would need to modify the workflows to be able to edit) so that you can indicate the original project for reporting purposes - especially for financial audits relating to sw capitalization that we have to show the stories/features that were worked on that have associated time and they all need to tie together.
3) You would have to go back and edit versions to make them unique to avoid conflicts.

Thus, keeping such issues in the original project is beneficial when its needed; but, the majority of individuals only want to deal with current work and backlogs - not issues that were closed 5-7 years ago.

You may already have this suggestion logged; but, I have the same issue when trying to search your own db's - I get so many results, I give up looking and log what I need - that is the same thing that happens to us and creates unnecessary noise and confusion.