I am a scientist by education and profession.
By the way you all have been discussing heat treatment, it is neither science nor art. It is a practice based in metallurgical science, much in the same way that the field of medicine is a practice of the biological sciences. Medicine, applied traditionally, is not a science.
When the chemical composition of a steel is known and heat is applied in a controlled setting, the metallurgical sciences have, through hypothetical-deductive processes, identified how that steel will react to certain amounts of applied energy, rapid changes in energy reduction (quenching), etc. When the variables are known in the system, the system behaves predictably. Science is the process that develops theory, which is the paradigm by which we make predictions about how steel will react, when the set of conditions of the system is known. If a theory is upheld long enough that we are pretty comfortable with it we may begin to refer to it as a law, or more generally, knowledge. Occasionally, new tests of a theory will call it into question, prove it wrong, and all the evidence developed up to that point must then be redirected toward a new, competing theory--a paradigm shift. (Hopefully we have all made the paradigm shift away from edge packing).
As heat treat practitioners, we have varying degrees of understanding of the theories that have been developed--through science--about how steel should react throughout the heat treat process. (What do you call a person that graduates at the bottom of his class from medical school? You call him doctor!) We can apply those theories in an uncontrolled setting and achieve some sort of outcome--but it is not science. When you start dabbling with steels of unknown (or at best, uncertain) chemical properties, apply heat in an uncontrolled setting, quench in a medium of uncertain capacity to draw heat from the steel, and then apply a modicum of heat to draw temper--then the theory that has already been developed is your best friend. You'll probably get within your mark 75% of the time, maybe better. And with practice, once a HT practitioner becomes accustomed to a particular accumulation of tools he or she might be able to improve on that percentage (though you will never REALLY know, will you? if you don't have the ability to control the circumstances by which you reach an outcome, you probably cannot critically evaluate the outcome itself.) Regardless, this is not science it is systematic guessing.
Loosely applying metallurgical theory, a practitioner might be able to identify a decent heat treat protocol for an unknown steel which, when applied along with a particular blade geometry, will perform to the makers' satisfaction, much the same way in that a surgeon might treat an aneurysm along a bifurcated descending aorta (not a normal anatomical condition, but bifurcated descending aortas are known to occur). But has anything new been learned about metallurgical science? Well--no. At best you've only supported the current paradigm and confirmed that the unknown steel does have a sufficient amount of carbon and likely a cocktail of a few other elements that may or may not work in your favor. At worst. . . at worst hopefully the practitioner recognizes that the mystery metal will not harden but might be suitable as stock for guards/bolsters.
Is the practice of heat treatment an art? Not necessarily. A HT practitioner might become very skilled at the HT process, and may even produce a quite serviceable knife, but one without any aesthetically pleasing qualities. Such a practitioner might be considered more a craftsman, than an artist.