Planned features, such as searching for an existing reference by content, will require us to implement some http://en.wikipedia.org/wiki/Unicode_equivalence .
We will probably want to use NFKD ("Characters are decomposed by compatibility, and multiple combining characters are arranged in a specific order.") to catch cases such as 'ﬀ' === 'ff', and we will probably want to strip combining characters (i.e. all accents), so that 'Amelie' === 'Amélie'.
https://github.com/walling/unorm looks like a good library for the job. We may want to fork it into UnicodeJS.