Here are a few off the top of my head:
Donald Fagen
Amazing in the studio and perhaps even better live. This guy is the man.
Rolling Stones
Disappointingly good, not great, in person. Sounded terrible in Athens for Bridges to Babylon tour.
Patricia Barber
Circa 1986 live at the original gold star sardine bar. Just amazing. Totally, hypnotically mesmerizing. Recordings very good too, but before she got a little kooky, Ms Barber was a real musician's musician.
Diana Krall
At Radio City Music Hall, a bit amplified and very schticky, head tossingly sex kitteny commercial. In the studio much better, I think.
James Brown
BB King Blues Bar, Times Square, circa 2000. Much better live. Usually badly recorded.
Blossom Dearie
Better live in Hell's Kitchen, circa 2001, but not bad on CD.
Lionel Ritchie
Private party, circa 2003, Southhampton NY. WOW -- I didnt think I even liked Lionel Ritchie and this guy was so cool I couldnt even believe it!!!!!!! One drummer, one bass, small amps and the keyboard.
Henry Rollins
Brixton Academy, circa 1998.
WOW! Now this is a tough call. Ear splittingly loud concert in small venue, but still sounded great. His recordings (at least the one with "Liar" are very very good, almost worthy of an audiophile recommendation.)
Beethoven Symphonies The Proms Royal Albert Hall
Forget which orchestra. Once in a while, a really good simply miked production will sound better than "live" but live usually wins here.
Michael McDonald BB King Blues Bar, Circa 2002
This didnt sound great, despite the smallish venue and intimacy with the small audience, the grey hair and the belly. But no doubt, this guy still has something special. I would say this depends, subject to the performance and the recording. Living on the Fault line is a good track for an audiophile.
Preliminary Conclusions
I think the moral of the story for me is that the bigger and more commercial the concert, the more likely it is to sound like crap and disappoint, even if we are dealing with charismatic energetic performers.
Disclaimer
Of course, substance abuse will alter the results of any survey.