An artist and an AI, when given the same prompt, will produce similar outputs. However, an artist replicates it in strokes, while AI replicates it in pixels. AI can create art, because art is in the eye of the observer, but its different than a human creating art.

Is there a turing test for art, and what's the detection quota?

I think any clear definition will either positively identify lots of AI works as art (along collections of random junk), or deny the qualifier to lots of supposed artworks from human artists.

Coming from theater, I agree it is about "conveying a meaning beneath the surface". Having studied computer science, I note that is very much not in a strict sense, but very vague. It seems to be a feature, not a bug, that everyone in the audience can see something different.

I think you can pretty much present random nonsense, and someone will still find it brilliant and inspiring, and a lot more people will tell you what patterns they saw, and of what it reminded them. The meaning is created in the minds of the observers, even if the creator explicitly did not put another, or any, meaning into the "art".

this is an interesting one cause it feels like a mobile philosophical goalpost, what would classify as 'feeling enough' for gyou?

Definitively the AI is able to understand the meaning behind a prompt and expand on it, before I've asked it for a picture of a cartoon cat and instinctivly it put a ruler beside it to show it was only a couple cm across

It certainly is a very efficient form of this compared to what were used to, cutting about as many corners as you can - but then again it still produces the output, and what other goalposts can we reliably argue for?