Regarding 'Lu' (Is, Am, Being)

[omängum fra'uti]

But my point is you can't just have two completely unmarked nouns for any verb, including lu.

Why not?  For a copular verb especially, such a pattern is common all over the world.  I've seen nothing in the Corpus nor the interviews with Frommer which hints at such a restriction.

But have you seen anything in the corpus or interviews where he used two un-marked nouns with lu?  The absence of proof is not proof of absence.  As long as we don't know for sure, any time you use it you're being speculative.  Admittedly a lot of things we've been using are speculative, but that's because we still have large gaps in knowledge.

However given the free nature of Na'vi I'd be inclined to side with "you can't have two unmarked nouns" because there's nothing to say which nounsmeans what - that's what the marking is for.

Anyway, I'd be surprised if there's anything that you'd want to express with two unmarked nouns which couldn't be expressed differently in Na'vi.

Edit:
Ok there is the one you already mentioned...

Tawsìp ngeyä lu sngeltseng
starship you-POS be (trash? waste? garbage?)-place

I'd give the same example without using two unmarked nouns there, but I'm not sure how or if combining nouns with adposition for the dative case would work.  Perhaps, then, lu can be used to say all subjects of the clause equate, in which case it would indeed be acting more like a copula than just a straight verb.

[wm.annis]

But have you seen anything in the corpus or interviews where he used two un-marked nouns with lu?

Sure.  The "your space ship is a garbage dump" example.  Neither the subject nor the predicate has any case marking.

The absence of proof is not proof of absence.  As long as we don't know for sure, any time you use it you're being speculative. 

But I do know that Frommer has said in multiple interviews that everything that occurs in Na'vi occurs in some human language, just not joined together in the same language.  There is nothing magical about a possessive (ngeyä) that would naturally lead one to believe lu has this restriction you're presenting.  If there were such a restriction, I'd expect to see signs in the garbage dump sentence.  Classical Arabic does operate this way — the "garbage dump" would be in a different case.

However given the free nature of Na'vi I'd be inclined to side with "you can't have two unmarked nouns" because there's nothing to say which nouns means what - that's what the marking is for.

But you don't have to have such markings always with a copular verb.  All over this planet people get by with copulas that enforce no formal distinction between the subject and the predicate.  Hell, the Russians don't even use a copular verb when it's in the present tense ("Yuri doctor").  Why does this cause them no ambiguities?  Context.  This is always hard when discussing grammar formally — we take up a sentence completely ripped from its context. Frankly, taking a sentence out of context changes the meaning of that sentence almost as much as pulling words out of it.  In an actual conversation, there would be no ambiguity between an unmarked subject and an unmarked predicate joined by lu, because everyone in the conversation knows what's going on, what's being discussed, and could assign subject and predicate without difficulty.