Esperanto isn't too much of a failure; at least, it's the most successful a conlang has been -- a couple dozen have learned Elvish or Klingon; a couple hundred thousand have learned Esperanto.

And yes, there has been a long and somewhat-pointless philosophical struggle by people who take language as more important, and those who take technology as more important. I remember discussing this with a Swedish man who insisted that we thought of everything in language, and I was insisting instead that I think in thought, which does not necessarily map to language. What finally seemed to make him uncertain about his own hard-line position was to just say, "look, I sometimes have the experience that I don't know how to put my thoughts in words. And I at least appear to be monolingual. It sounds like you're saying that this should be literally impossible. My thoughts are always in English, how could I ever have problems encoding them into English?"

Stephen Pinker's works have indirectly shown me one useful thing about these philosophical discussions: it helps to have an idea of Gödel/Turing-completeness[1] in your philosophy of language. The problem with the radical biological innatists[2] and the radical linguistic relativists[3] is that it seems hard to describe how languages evolve under either account. If our languages and proto-languages are in some sense Gödel-complete and can describe anything which is describable, then linguistic evolution just consists of building new language constructs which can more efficiently describe the new technologies -- and these in turn help us in thinking about our new technologies and making even newer technologies. You can also get a good distance in understanding how children acquire language; it's fundamentally similar to the bootstrapping process which gets a computer from being offline to running a modern operating system; a small set of built-in intuitions can construct a model for the entire rest of any logical system.

[1] Gödel's completeness theorem, as distinct from his more famous incompleteness theorem, says that a couple basic laws of logic are all you ever need, and any other logical system can be emulated by just constructing new objects whenever the axioms of the basic system call for them.

[2] something like, "all the concepts which you find meaningful are somehow indirectly encoded in the genome, your words name just those concepts meaningful to you."

[3] something like, "if your language doesn't have a word for X, then you just cannot think about X, full stop."