This thread was lively.
Sean Olive and Harman’s approach are founded in the scientific method, there is no doubt about that.
However, I accept their sample sizes could be inadequate to establish universal preferences.
10 to 100 subjects may well be insufficient to draw absolute conclusions about what millions of people prefer.
That being said, no one will ever fund such studies with 1 million subjects. There’s no money in it and it’s not practical.
Does it mean such research has no value?
No, it doesn’t.
Human physiology exists in a relatively narrow bell curve.
Such research and measurements can tell you if something has a chance of being liked by many people; it can also tell you if something is likely to be utterly broken “by design”.
Yet, it will never be hard to find a product with 100,000 5-star reviews on Amazon that doesn’t match Harman’s targets.
So, in conclusion, Harman’s targets may perfectly match your preferences, in which case the graphs Amir gives you for free are gold.
Or, Harman’s targets may sound “not quite right” to you, in which case Amir’s graphs will also be useful.