Why are older tubes so much sought after

Mil Spec components were usually required to be much better than necessary. That was before cost control became important. Nowadays most military equipment uses commercial hardware whenever possible, which greatly reduces cost, improves (yes) reliability and performance. Also, we have learned to make electronics whose performance is derived from the basic circuit design rather than from reliance on tightly spec'd components and trim procedures.

I am surprised that new tubes are less accurately made. A tube would seem made to order for precision automated fabrication.

Well designed electronics OUGHT TO BE insensitive to minor variations of components, including tubes. Perhaps this design objective is not met in some audio equipment. All things equal, an old tube should be inferior to a new one because of gas infiltration over time. Maybe gassy tubes sound good...more of that tube sound.

Larryi...The design of electronics so as to not rely on precision parts or trims is a well established practice in areas other than audio. In my area of work (missile guidance systems) we were forced to adopt this philosophy, quite apart from cost benefits, because requisite performance was not achievable simply by using more and more exotic parts. There were howls of protest from the circuit designers, but once they put their minds to it things have worked out well.

One obvious example from audio is automatic tube bias adjustment.

As for evidence about gassy tubes, I simply submit that gas leakage through glass is a function of time so old tubes can be expected to have picked up some gas over the decades. Of course a new tube might have gas from the get-go. Some large tubes, notibly CRTs, have a "getter" which captures gas when the tube is operated, and failure to turn on such a tube, or at least the filiament, for a few years can result in catistrophic failure.

NOS tubes are compared with new production by listening to them. It seems to me that comparitive measurements of what the tubes do to the signal waveform could lead to an understanding of what needs to be changed in new production tubes to make them sound like the old ones which people prefer. The laws of physics haven't changed, so there is no reason that tubes made today can't be superior.

So I guess that the one thing that everyone agrees about is that tubes "color" (alter) the sound: opinions vary regarding the best color.

But I thought the objective is to reproduce sound with "High Fidelity", which means without alteration. There really is nothing wrong with altering the sound if that's what you like, but such a system should be regarded as a musical instrument rather than a reproducing instrument.
.

Newbee...I was around in those days. Our tube equipment hummed, hissed, and produced so much IM distortion that one percent harmonic distortion was not a problem! Our goal was to improve these measurements, and I guess we were successful except for the IM distortion which still runs around one percent in many (most?) tube amps.

By the way, how would you describe "grain" and what would cause it in a ss amp?