As there are many text strings to translate, I guess the best is copy the English Version, not translate the LS standard strings and just call it "Inuktitut". All the standard texts will appear in English. But the ones that actually appear to the user aren't that many, so either you leave them in English or you translate just the ones that you need for your survey (e.g. error messages, instructions, etc. that really appear for you in your questionnaire).
Because if you want to translate the whole language file, this will take forever and many strings are actually for the survey backend, that you don't really need in "Inuktitut".
I think we should have two language files, one for the survey strings and one for the backend. I don't think we need the backend in that many languages, but we will need the front end in as many languages as possible.
I made a feature request:
bugs.limesurvey.org/view.php?id=15245
Might never make it into the core, but I think it would help if we could just translate those strings that are seen by respondents, especially for exotic languages. The work is a lot less and people can run surveys in more exotic languages easier. Because translating only those strings is also feasible for only one survey, now I guess few will commit to translate the whole Limesurvey backend as well, just because they need a survey in a rare language.