Why recordings made before 1965 sound better.

Part of it was that stereo lps were relatively new and still a novelty and records were often marketed and sold based on sound quality.  It was the golden age of vinyl. Then the novelty wore off and results became a lot more mixed.

Sns,

You raise a fair point. My feeling, though, is that the tubes produced a warmer sound than solid state, whereas the spatial element in recordings was compromised by Dolby interfering with the natural overtones of the music. By compressing and replacing the high frequencies the result is an artificial recreation of the natural sound.The openness of pre Dolby recordings is witness to that. It’s simply more real.

When did the vinyl become thinner? Hard to believe that didn't impact the sound. 

Neither Amos nor Andy.

it is a common misperception that digital noise reduction [sonic solutions NONOISE, CEDAR et al] per se,  "harms the music" - it ONLY does so when it is misapplied by a cloth-eared ham-handed audio restoration technician. having used these tools for 3+ decades now i can tell you that they are godsends for musical enjoyment of crackly hissy rumbly old phonograph recordings esp. those on 78. this rant out of the way, i can say that it is the ART of recording and not just the TOOLS of recording that matters more in terms of sound quality. i've heard well-done dolby A recordings that had just as much "air" as ones made a few years before dolby A came on the scene.  ya just gotta do the job right in the first place. 

Early stereo recordings of orchestral material might sound better because a few of the best recording engineers of that era were simply trying to capture what the conductor and musicians were creating within the great sounding venues of that time as opposed to "creating" something of their own.  The more simplistic miking techniques they employed went a long way to capturing the spatial quality of the performance.  Most of the early stereo recordings were mastered on triple track 1/2" recorders like the Ampex 300 or RCA RT-21 machines and were very carefully mixed down to the 2 channel sub-masters used to cut the LP's.  Tape hiss and speed variations were audible, but not objectionable with most performances.  In an effort to minimize these issues, 35mm magnetic film recorders were modified to record three 200 mil tracks which were over twice the track width of those on tape, increasing the S/N ratio significantly. The film moved at 18" per second by means of a sprocket driven drive system that virtually eliminated wow and flutter.  The media for that technique was prohibitively expensive, but LP's cut from those masters are pretty impressive.  I'll avoid the tube vs. solid state debate, but maintain that it was mostly miking technique that made the difference.  2 or 3 well placed mics were all that, in most instances, were needed to do the job.  Every additional mic added could potentially spoil the Vichyssoise and guys like Lewis Layton, Bob Fine and Ken Wilkinson knew this well.   The maestro and musicians did the best job of balancing everything. The job of the recording engineer was simply to capture the performance.