Ive stayed out of this thread. The subject is rather too big and complex for the sort of reasonably brief discssusions preferred here. I get challenged from time to time for my wordy posts, so have hung back... But let me try.
I spent 3 years in the early '80s as Quality Manager on the staff of the VP of manufacturing, for a company that made stuff complimentary to the businesses we discuss here. That business was arguably as technically complex, if in a different way, and perhaps more diverse than binos, scopes etc. In those days I was one of many US Quality Mgt execs schlepping to Japan and asking them what, how they were doing what they were doing. I sat in one meeting with the heads of a rather large industrial conglomerate where they said they were surprised I/we were coming to Japan and asking these questions as we had taught them! Indeed it was Edwards Deming, Joe Juran, Frank Gryna who had been consultants to post WWII Japan and helped rebuild their industrial base. Back then they were back in the US teaching, writing, lecturing, Their work, then called Total Quality Management, was standing on the shoulders of Statistical Process Control, that formed the basis for Just in Time or the Toyota system of manufacturing. I heard some of their lectures, read their work, hired a super qualified consultant to guide my uninformed, amateur effort. Thankfully I had an amazingly tolerant boss.
Relevant to this discussion, I thought when posted it would go down the road of how things get made, how quality is controlled, the role of sampling, statistics, the stuff that ensures the quality level of products to be shipped. I was surprised but in hindsight should not have been, when it took the tack down this road of piece to piece, optical quality variation.
In theory, as Henry describes, there is a perfect or near perfect example of a thing, and a way to ensure whether one has acquired that or not (star testing, etc). And as he wrote folks with 2020, 2015 or better eyesight can appreciate such things. The problem for a manufacturer, is where to draw the line, how good is good enough? How to practically establish standards that can be measured in a meaningful, efficient way as a compliment to the production process.
I suspect many know of that old notion, "The perfect is the enemy of the good." To produce an item that satisfies as many people as possible at a price they're willing to pay that provides an acceptable return to the owners, requires compromise. I suspect its fair to argue the better companies, know this well. The success of their products and their continuing existence suggests they take pretty good care of most of us most of the time...
I agree it takes a lab, with sophisticated test equipment, and super careful controlled procedures to find the best of a thing. Knowing that level of performance informs all kinds of useful things. I also agree there are applications where that degree of perfection is warranted in use. Those that have asked above, "if we can't see it, especially if using my binos the way I do, should I care?" are onto the dilemma faced by the makers.