How close to the real thing?

Somewhere in the miasma of the OP's question is another beguiling paradox, namely:

What percent of the time has live, unamplified music ever been mistaken by anyone as being other than "real"?

If the answer to this question is close to "zero" (as in "zero" percent of the time I have mistaken live music for being recorded ) then, how is it or, rather IS it possible for the OP's question to be other than the reciprocal?

If OP's friend is right, then it just might be that 5% sound pretty darn good, but I suspect "5%" is just hyperbole.

I only disagree with the 5% part. I think that should be more like 90% and mostly constrained by the room size. A properly set up system in an auditorium could get much closer to the real thing.

As a trumpeter, I've seldom felt that I heard a recorded trumpet sound as live. There were a few direct-to-disk exceptions in the analog days and some reel-to-reel recordings got close, but most lacked some key overtones and dynamic range. Now, with DSD and other hi-rez recording techniques, I'm hearing the real thing more and more.

Interesting thread. I agree with Mapman, Atmasphere, Shadorne, and Bryon. Certainly anyone who thinks that they could possibly reproduce the sound of a full symphony orchestra in their room and have it sound remotely close to how it does live is fooling themselves. Though I know he doesn't particularly care for concluding it, Bryon is correct in saying that there is no single answer to any of these questions.

One interesting thing, Bryon - you mention that "The greater your imagination, the easier it is to fill in what’s missing during recorded playback". This does not necessarily even require imagination - if one was present at the live event that one is listening to a recording of, for instance. I frequently listen to the archival recordings made by my orchestra (I am on the broadcast committee), and I am constantly surprised by what is NOT picked up by the mikes, much more so than I am at what is picked up.

I do know what you mean, though. In fact, this imagination can be a liability for musicians that are trying to judge recordings they have made of themselves for audition purposes, for example. Also, sometimes it is difficult to tell when listening to such a recording whether the sound you don't like is your fault or the fault of the recording (for example, you think your sound is too bright). This just happened to a colleague of mine, who had to come to me to ask my opinion - this issue can cloud one's judgement very easily, messing with your head for no reason sometimes.

So far JAX2 has been the closest as the correct answer is...... 65%. Still lots of room for improvement.
So, lets keep at it.