A titles-only search result, with zero extra post processing (no page ranking, no snippet, no highlighting, no page info, no parenthetical saying it matched the category or the redirect or the section) seems like an attractive, win-win feature.

The old search depended on page ranking, offering only intitle and incategory for parameters, and pouring out myriad results. Now we have many new Search features which eschew page-ranking, enabling us to finally target a specific set of pages with precision for the purpose of definitely handling each and every one.

Providing this would encourage:

• post-processing those pages, targeting those titles in AWB
• presenting those titles elsewhere, adding markup for wikilinks or checklists
• comparing one query to another. A quick diff or cmp helps learn Search, diagnose Search, develop Search documentation

Other attractions are

• five times more information per page
• clean, relevant, uncluttered results compete with SpecialPages in general, PrefixIndex, WhatLinksHere, Categories in particular
• an intitle search with titles only

I realize that the end-user who are post-processing can do it themselves.
But currently there is no obvious way to grep titles from Search results.
(Personally I have javaScript add an edit tag to each title, and can grep on that.)

Currently the end user must do the search, go to the bottom of the page and select a higher number of results (than the twenty), have the search done again, and then undo 1) the snippet 2) the document size and date (or category member counts), and 3) the parenthetical saying whether it matched in a category box, or matched in a redirect.