Let me suggest a somewhat broader point of view. We are mostly thinking about "the scientific method" and "the scientific process" in terms of observations, proofs, theories and so on. However, as it is clear from other discussions here, there is a great amount of "soft-skillish" element involved, as with any activity that concerns communication.
As such, "a paper" is a way to communicate / spread our ideas, and if some theory of ours doesn't take roots, it means that we aren't very efficient communicators. Thus, even if a theory is eventually published in a respected journal, still a bumpy publication road may indicate that our way of transmitting ideas isn't particularly appealing.
Therefore, there is certain psychology involved, too. For the sake of the argument, suppose the theory I am pushing forward is sound. Even in this case, people are wary of "grand theories", where they first have to accept certain premises, and then see what kind of great building is being constructed in front of them.
I'd say that a more pragmatic approach would be to attack specific individual problems and show that you can solve them. Yes, it might be based on a certain (unknown or unaccepted) theory, but why bringing it in right away? Here is a problem, here is my solution, answer obtained, knowledge increased, everyone happy. After several success stories of this kind one might try to unify them all under a certain common umbrella of a "proposal" or "presumption" and let it to take root.
In other words, we should remember that we write for the reader. Why readers should care about our theories or take time to ponder whether they are sound or not? Readers seek something for themselves, and the best way to push our agenda is to show how it aligns with possible goals of our readers, and give the readers tools for their tasks in bite-size chunks, if possible, without overwhelming them with grand structures.