We have had this discussion before - sorry I haven't the time now to find the old thread, but I will say again what I said then:
If we use a word which means nothing at all to the listener, we have failed to communicate. If we use, instead, a word which is close to the meaning we want and is
understood by the listener, we have communicated at least part of our meaning.
Examples: (1) dukkha / suffering; (2) sesquipedalian /long
(Note that it doesn't matter whether the obscure word is Pali or English.)
People learn the meaning of an unfamiliar word by having it explained in words they already understand, and then (usually) by seeing it often and in a variety of contexts.
Again, it doesn't matter whether the new word is Pali or English.
If a foreign word is used often enough in English, it becomes naturalised - an English word, in fact. That happens most often when English doesn't already have a word for the idea. Sometimes it retains its original meaning, sometimes the meaning changes a bit - but the meaning of every word changes gradually, so maybe that's not an issue.
Different communities and groups of English speakers have their own vocabularies. English-speaking Buddhists are well on the way to naturalising 'dukkha', 'nibbana', 'sutta' and other Pali terms (and their Sanskrit equivalents). You could even say that knowing those words is a sign of membership of the group. In that case, using them within the group is not problematic, but using them to non-members will result in communication failure.
It all becomes a matter of using the most appropriate word for the situation, then, doesn't it?