Looking into this more deeply, it seems like the best way to do this is probably to import everything with a specific tag, then in the vocabulary section of LingQ filter on that tag and repeatedly mark everything as known (unfortunately have to do this many times as the maximum you can select in a given view is 200 words).
If there is a better way to do this, please let me know!
It also looks like importing a word that you already know will overwrite the knowledge status for that word, which will make it very difficult to import new known words as I work through them in an independent SRS system? Is this a correct understanding of the situation?
Yes, that sounds like a good solution. Other than that, you can import Vocabulary through the CSV file, also assign a specific tag, and then later change their status.