As described in anemail exchange on the ArchivesSpace member list in 2018, the repository details (institution name, address, etc) are indexed in both the staff interface and public interface searches, which can be a major usability issue. In Joshua Shaw's example, his repository's address coincides with a particular person who has material in the archives, making searching for this person's historical records extremely challenging. For us (and probably many other repositories), we hold records of our local area, so searching Lansing and East Lansing is pretty useless because of this issue.
From the discussion, it sounds like a lot of thought and work has already been done around the issue, but it seems to still persist. I wonder whether there's a way to exclude the repository information and citation info from the search results? An argument for doing this, beyond weighting hits in other fields, might be if people try to do data visualization/analysis on the collections, which would still be skewed if those fields remain indexed. Thanks for your consideration!