As an engineer I prefer to use the word "optimum", becasue there always is a optimum solutions to a problem and a given set of contraints. The "optimum" may not be the "best".

As an audio engineer the thing I have always liked about the "older" stuff is its simplicity. But, also as an engineer, we have come a long way in our understanding of audio and the materials and tools that we use in our designs. As engineers we would be doing a pretty poor job if somehow we hadn't improved on the old. I consider myself a good engineer so I think that I can improve on the old. But I always try and do it as simply as possible. Don't usematerials that don't matter. Don't add components that don't make a difference. Don't pay more for something if it isn't any better. In short - optimize the design.

In my expereince, the audio "optimum" is converging on some final principles. I won't belabor them all here, but a few are readily apparent. In electronics its minimize feedback. Among amplifier designers that I know (some of the best) this is pretty well standard thinking - tubes have always had this. In loudspeakers, it appears to be good polar control. The Harman guys and I certainly agree on this point. Room acoustics - well thats still in turmoil I'm affraid to say.

At any rate, I am partially agreing with you that as the constraints become less and our knowledge increases, the "optimum" solutions should (and will) converge on a finite set of answers. In principle this set can be as large as the number of degrees of freedom in the contraint set, but in reality, I think that it will be a whole lot less.