I merely gave an old example and my area of expertise is not statistics. There are many newer examples out there. It could be jelly beans in a jar.
The example with a group of people and an ox may have been skewed if farmers were there who were used to such things. In other examples random people were chosen and the means or median answers were quite close
Generally what we do is a group effort and most of the people involved have backgrounds in engineering, music, mathematics, computer science, even the medical field. There are practical purposes involved I will only get into in PM.
I was doing some blind testing with specific tubes on a small scale and several friends were curious if we could expand the testing, and how to go about that. It is not for everyone and it can be costly. If you have a better methodology by all means use it.
This is very specific but we did DAC testing in a similar manner. Once again we got to hear equipment that we could not walk into many showrooms and hear, and yes we did score them blind.
I am not one to say I hear this, and this equipment is better or worse so I like to get several people in the same situation and get multiple thoughts. When I designed the four amps friends said they liked them but I figured some were humoring me so I sent a sample on to folks on this site I have never met and got their reactions. It is just how I think.