steve can be a fun guy to listen to and some of his stories are, well, the kind of stories one might imagine working in a high end store in nyc back in the day.
he doesn't seem to be familiar with the scientific work that has emerged over the past 10-15 years or so that has found very strong correlation between certain aspects of sound reproduction and subjective listening preferences. it may be because he isn't aware of the research, doesn't have the background to interpret the statistical measures, that he has beliefs that at this point he isn't willing to change, or it could be something else.
in any case, when subjective listening preferences are on the order 90% predictable based on purely objective measurements, there is very little room for "all the other stuff" that steve is talking about. once "errors in the variables" are further considered, which arise from the fact that people may not reliably pick their "favorite" each time in a blind test, the subjective part that remains is likely down in the low single digits. also, there is an adaptation aspect. over time, "your" speakers will begin to sound more right as your brain makes its own eq adjustments. when presented with sound that is more accurate, the brain may initially find it sounds a bit "off", but over time that aspect will adjust as well. this further limits the statistical power of the tests, which also reduces the remaining portion available for subjective preferences not explained by objective measurements.
that's not to say that speakers have different presentations of sound and different types interact with rooms differently, they do. but just based on the sound characteristics alone, at this point, it is pretty easy to predict what most folks will think sounds good.
as a side bar, olive gets the award for the funniest conclusion in a scientific paper:
males prefer more bass than females. going out on a limb here, i don't think more research is needed on that one. :-)