Smart guy, Gould, but my conclusion is that his argument for recorded performances OVER live concerts is--to me, at least--overstated. Also, one might quibble with his personal definition of the word "performance."
"Performance" means, for most musicians and audiences, a "live" performance BEFORE AN AUDIENCE and IN A HALL or VENUE of some kind, with all the risks and all the advantages that a "performance" in this sense entails.
The word is also used in an auxiliary sense: a live performance before an audience heard indirectly, that is to say, on the radio, or internet, or streaming site, or CD.
But GOULD deviates from that usage, which is fine, except that he doesn't acknowledge the deviation. He calls his heavily edited recordings "performances," which, where the edits are HUGE, makes the recording less and less a "performance," in the usual sense.
Nothing wrong with editing, even tons of editing. But where and when the question is put: "Is that a recording of X "performing," the answer must be "not exactly," or "not completely." That answer is required so as not to confuse or mislead listeners. (I appreciate this distinction from my own experience, since everything I personally "record" has been, over many years, subject to repeated MIDI edits as I find things I don't like.)
The other respect in which Gould "overstates" his case is the claim that recorded music REALLY IS BETTER, which means OBJECTIVELY BETTER THAN live performance. Surely this issue boils down to personal preference, and that's it. Some folks prefer recordings, on the whole. Others prefer, overall, the live experience. And yet others, which I suspect may be most people, like BOTH more or less equally for all the obvious reasons, which don't need to be rehearsed here!