Today's case in point is this paper (hat tip to turbo-1 at the physics forums for finding it):
Is a classical Euclidean TOE reasonable?
Authors: A. Arbona
We analyze both the feasibility and reasonableness of a classical Euclidean Theory of Everything (TOE), which we understand as a TOE based on an Euclidean space and an absolute time over which deterministic models of particles and forces are built. The possible axiomatic complexity of a TOE in such a framework is considered and compared to the complexity of the assumptions underlying the Standard Model. Current approaches to relevant (for our purposes) reformulations of Special Relativity, General Relativity, inertia models and Quantum Theory are summarized, and links between some of these reformulations are exposed. A qualitative framework is suggested for a research program on a classical Euclidean TOE. Within this framework an underlying basis is suggested, in particular, for the Principle of Relativity and Principle of Equivalence. A model for gravity as an inertial phenomenon is proposed. Also, a basis for quantum indeterminacy and wave function collapse is suggested in the framework.
This is not discussing a mainstream theory. It contradicts the basic principals of general relativity, with both absolute coordinates and an ether approach to the vacuum. It seriously rethinks whether the speed of light speed limit of special relativity is absolute. It proposes a major rethink of some of the apparent paradoxes of quantum theory. And, it does it all in one paper. My point is not that everything physics has held so far is wrong and that Arbona is right. Arbona may not even personally believe the theories explored in this paper. My point is that radical departures from the leading paradigmn, when proposed in a reasonable manner with appropriate equations and discussion and appropriate citation to the relevant scholarly journal work that provides a foundation for the theories can and do reach very wide audiences whether they are published or not.
Plenty of scientists will read the abstract and move on. Different scientists have different interests. But, it takes only one or two serious scientists in the field with an interest in the topic to read the paper and discern whether it has promise, in order for a shot in the dark good ideas to snowball. If the paper is good, those scientists will refer to it and build upon it, and the development (with its originators well documented), will quickly spread across the scientific establishment. Most scientists have read dozens, if not hundreds, of these kinds of papers. At any high level physics lab, the collective momentary glances of all the scientists involved in an experiement provide a resevoir of ideas to consider, should anomolous results deviate from current mainstream theories. This doesn't happen very often, but it does happen, and provides a means for supposedly suppressed ideas to emerge into the mainstream if they are needed to explain reality.