Underlying the discussion of whether this or that tube is better than another is the notion that there is an objective standard and that each tube fits into a specific spot along a worst-to-best continuum. That is hardly the case. It is a matter of the designer or the user finding a personal fit. There are so many more choices in older tubes, it is not surprising that one can, if one searches diligently, find older tubes that better fit the bill.
Manufacturers have to use readily available tubes, meaning, in most cases, currently manufactured tubes. Why go through the considerable expense of selecting NOS tubes when there is a good chance those tubes will not be the customer's preference anyway? Also, for those not inclined to try different tubes, the manufacturer has to be reasonably assured that the same tube is available many years from now to service that equipment. Just by pure odds, it is unlikely that the particular tubes chosen by the manufacturer will fit any particular customers idea of an optimal choice.
I am sure that, wholly apart from sonic considerations, certain older tubes have because of their "exotic" and rare status (people actually collect tubes, such as 300As). But, i can assure you that I would be glad to use current 6sn7, 12ax7s, etc., if they happen to work well for me. I pay a lot for the EML meshplate 2a3s I run because, they outperform anything else I have heard. In other words, I am not paying for their rarity (they are currently being manufactured), but I am paying a premium for performance.