What Bob Carver did was very specific. He made the assertion that the only difference between high-end amplifiers was embodied in a reproducible transfer function… Basically that the nuances of sound that aficionados of given brands favoured and heard was not down to some necessarily very expensive secret sauce or magic (which the manufactures tended to promote as the reason) but something very simple that was able to be captured and was reproducible. However what he did was seat of the pants, and the amplifiers he marketed after the fact based on the challenge had a reputation of slipping out of alignment and sound unlike they were meant to. His challenge allowed him to wrap on of his own amplifiers in a network of stuff that he hand tweaked until he nulled out the sound difference. This nulling process is not the same as measuring it - he may only have had a very basic idea of the parameters of the particular emulating circuit he was tweaking, but he could tell when the two sides matched. He would be doing little more than ensuring that the input and output’s complex impedance was the same, and likely adding a differential pair that could be unbalanced, and maybe a single ended driver that could also be tweaked, enough to match the broad harmonic distortion characteristics of the emulated amplifier.
You can buy an argument with Bob’s work, there is a lot of nuance in what was done - but there is little doubt that he identified the single biggest contributors to differences in amplifier sound. Even he would not claim he could do the same with loudspeakers. Loudspeakers are a whole new can of worms, and the notion of a single frequency response is meaningless (despite what manufactures tend to promote), and the distortion mechanisms are time dependant, which makes life miserable for modelling.
Human taste is a very difficult thing. Something as complex a mess as the results of fermentation and ageing, with variations in the flavours of the original material, yeasts chosen and maturation should give a clue about the hugely complex mess of chemicals that end up in the final produce. A neat thing to get is a wine tasting odour training kit. These contain samples of specific chemicals that yield the basic set of scents you will find in wines - including off smells. But they won’t allow you to go out and reproduce a wine. It is still far too complex a mess. The OP’s article concentrates on whiskey. (But not whisky ) The raw whisk(e)y is likely more more easy produce to grasp - it is more limited by what will come over in the distillation process. But maturation is another question. However, being is the US, the article is probably only talking about Bourbon. So nothing that has actually been cask aged. That sets the goalposts awfully close.
OTOH, the application of science, in a much broader manner has revolutionised things like winemaking. Even the cheapest lowliest wine should be perfectly palatable and drinkable. Building a very high end wine is still a matter of skill, but it isn’t a hit and miss affair - winemakers know exactly what they are doing and what to expect as the result.The rules of the game restrict their options - although this varies from country to country.
Perhaps he OP’s question can be narrowed from “what has science done for us in improving things” to examples where a singular application of science has enabled a cheap thing to be tweaked to be just as good as an existing expensive thing.
One area that comes to mind is metalurgy. Tiny tweaks to the composition, or fabrication techniques (ie heat treatments) can wildly improve the capability of an alloy.