About performances and recordings -- chances are, it's true that some players of yesteryear might not seem up to today's standards. I have a copy (in English translation) of Schweitzer's autobiography, in which he has a whole chapter devoted to organs and playing. He's somewhat opinionated, but he makes the remark that Bach is "played altogether too fast".
I sometimes wonder if that might not be the key. With modern organs, do many organists tend to play at faster tempi than was once the norm? Moreover, acoustics in a church or hall can make a difference. It would seem that, in a place with a long reverberation, fast tempi, especially with baroque-style music or anything with toccata-like passages would come across as just a lot of noise with little attention to individual 'voices'.
I have a complete set of Beethoven's symphonies done by the Chicago Symphony under Sir George Solti. In the booklet he makes quite a to-do about faster tempi than were once "standard". But supposedly Ludwig left 'metronome markings' for them. If true, why did it take conductors 150 years to discover them? So some of this is a matter of opinion, I suppose.
It seems to be documented that Widor was highly reluctant to do recordings -- partly because of his age by then, and probably also because in those days recording equipment could do little justice to the situation. It seems to have been when "high fidelity" recording came along in the 1950's that the organ suddenly became the test of recording.
All in all, there is probably some truth to the idea that things were different a few generations back.