One of the best ways to train the mind in terms of critiquing science, aside from philosophy and critical thinking, is poetry. The reason is simple: poets are experts on metaphor. Scientific explanation, in the sense that it attempts to describe an (actually) impossible potential perspective (how things are for themselves and each other, rather than for us) inherently involves an as-if which takes the form of metaphor.
An example is the theory of radiation pressure maintaining the size of a star (preventing infinite gravitational collapse). The theory takes a metaphor (in this case “surface” of a star) and treats it as literally true. The problem with this is illustrated by the following:
Gases (which all stars are composed of) don’t have a well defined boundary that you could call a surface. Surface tension is caused by molecules in the bulk experiencing inter-molecular force in all directions while molecules at the surface only feel that force in one hemisphere. This causes a higher density layer of molecules near the surface in comparison to the bulk. Only an entity with a well defined boundary that has at minimum the type of surface tension found in liquids can create the resistance to radiation necessary for “pressure” to affect its size.
As a result the theory of how stars are formed by gravitational collapse of accumulations of hydrogen gas falls apart, since there is no available counter-force that would stop the process, it would accelerate until many of the base particles were ripped apart and therefore, in any meaningful sense, would no longer exist (since they would have no effect on anything). The last point is an implication of quantum mechanics that black hole theorists appear to have completely missed, and points to a more reasonable notion. The process supposed to lead to black holes would generate sufficient internal gravitational force that most of its mass would be ripped apart to the level of massless, force-less particles that cannot be detected. At a certain point the amount of mass-generating (and therefore gravitation-generating) matter would simply be too small to continue the gravitational collapse (at the point that particles are close enough to allow the weak force to counter the lowered gravitational force), giving a range of star sizes dependent on the initial size of the gas cloud and the physical proximity at which the weak force balances gravitation. All possible stars (in this phase space) would have to lie within which range.