I like the open peer review approach you are taking with this work. It's radical to let the community comment on the work as it's being drafted.
It's very much an orthodox memetics approach to attempt an answer at the "what is a meme?" question by asking what things memes are like, and then considering what other things these meme-like things are like. I think it's extremely telling that one of Dawkins' primary motivations for introducing the meme concept was to clarify the gene concept, which by no means is totally clear nearly half a century since The Selfish Gene. And so, if memes are like languages, and languages are like art, perhaps we can indirectly examine the nature of memes through what we already know about art and language.
@jayaoyama is currently working on his thesis about the ontology and epistemology of memes, and one of his key references is Wollheim, whose Art and Its Objects presents (I think) one of the most compelling theories of art ever.
Wollheim suggests that one might identify objects which are artworks without relying on a definition of art: we can identify art by first picking out certain objects as primary works of art, and then use them to identify derivative works of art by using some rule or another. I don't think this is how we typically identify and experience works of art; but this is typical of the way that we identify and experience works of meme. And this bears resemblance to a Quinean theory of language acquisition in which (pre-linguistic) meaning is facilitated via ostension (e.g. learning the word "water" through pointing at water and associating it with the utterance "water"). I wonder how well this model fits with how memes are experienced.
I know that there has been a huge amount of debate and criticism branching out from Quine's account, but I don't know the details. @Conner has written his Master's and doctoral theses on memes from a linguistic perspective, although he doesn't "suggest that memes are a language unto themselves". One of his main themes was ostension, so he will know a lot more about this.
I think that makes sense. It reminds me of how digital living documents and sites (e.g. wikis, "digital gardens") are sometimes described. Are social media platforms also "heterogeneous, unplanned, additive, experimental, fragmented manifestos, always incidental, always in continuous construction"? What about software, which can be patched and updated? Inasmuch as we consider memes as recursively identified by prior exemplars of memes (as when we apply Wollheim's theory of art to memes) and posting memes (or any other digital content for that matter) seems to involve some ostensive dimension that goes beyond making stuff visible to other users, I would wager that you are right to consider memes as constantly "making sense of memes".