I've generally always been in the dark ages.........
When I was in analog, I did try that "keeping a few takes and deciding which one" thing but I was limited for tracks and so that forced me to adopt, pretty early on in my recording life, the method I do now ~ get it all done in one go ! If mistakes are made, I'll just start the particular take over or, if it has been good up to the mistake, keep that and start from that point. Poor man's punching in.
When I was looking into 'going diji' back in the early 2000s, one of the big selling points of computers and digital recording was that "X amount of virtual tracks so you can compare takes". Even back then in my ignorance, I ignored all that because it meant and means nothing to me. I often record in sections so my SOP {that acronym is in honour of Miroslav} means that I'll end up with sections of, say, a vocal anyway {eg, verses, chorus} so then I just put them together {poor man's comping}. With my super long tracks, gone are the days when I would do the take in one go. When you have a 22 minute piece with different bits and time changes etc, sections is the way to go.
You know, when I used to read about all those bands that would do multiple takes and keep them then use take 2, I often wondered why. I tend to feel that if you know a take is good, go with it !