How close to the real thing?

I think the most interesting aspect to the original poster's question is the fact that it forms the logical basis to objectively improve our art - that is, "Does it sound real?" is actually a falsifiable question. The answer is definitely highly dependent on an unfathomably large number of personal differences in perceptional acuity . . . but one can always ask oneself this question and come up with a relevant answer. So the main thing we *should* be striving for in the technical studies of audio (i.e. measurement and controlled listening tests) is repeatability.

For reading on some of the logical basis of falsifiability, I'd recommend starting with the work of Karl Popper, or at least his Wikipedia page:
http://en.wikipedia.org/wiki/Karl_Popper
http://en.wikipedia.org/wiki/Falsifiability

Under the logic of falsifiability, it takes just ONE instance of having experienced reproduced sound as being real . . . to validate the conditional plausibility of all the technical and methodological basis for its reproduction. Not that the reverse isn't true as well, but if "reproduction" as an abstract is the goal, then at least for me its occurrences are much more noteworthy than its absences.

Here's the problem I have with setting up camp in any of audio's little belief-systems (i.e. tubes/transistors, cones/panels, feedback/no-feedback, analog/digital, etc.). If I'm truly honest with myself, I have at times been dumbfounded by amazing experiences from really crappy stuff, and also underwhelmed by my experiences with some really beautiful machinery.

And since audio is really a trivial pursuit (no lives at stake, no big "origins of the universe" questions), I view the the craft of audio thusly (my quote):

To examine a wide variety of data from both measurement equipment and listening experiences, and establish correlations that logically fit as much of the data as possible . . . then assume causality based on well-established principles of physical science.

This is why I think audio is so interesting in general -- we get to explore a truly genuine frontier, consisting of abstract and profound questions of the limits of human perception in sound and music. We also get to explore many aspects of human psychology in the listening process, and sociology in the marketplace and hobbyist communities. And really, the assemblages metal, wood, plastic, paper, and whatnot that we fashion to this pursuit are not really so profound . . . after all, it's all going to be in a landfill in a couple hundered years at the very most.

Atmasphere, here's the main reason why I'm so routinely perplexed by your dogged determination to proselytize the whole anti-feedback thing . . . What do you do when you have an amazing experience listening to music through a system, and THEN learn that it's a solid-state amplifier using a pile of global NFB? Because it's happened for me often enough that I can't imagine it hasn't happened at least a few times to anybody that's made a career in audio.

Well, the whole negative feedback issue has been beaten to death pretty thoroughly already in other threads.

My conclusion is that it is just one of many design and execution factors that affect results depending on how well it is executed.

Having heard a lot of gear over the years, my ears tell me that good engineers working for companies making good products have a much better handle on all such things that matter these days than they did 30 years ago.

Hi Mapman,
I was speaking to an old friend of mine who built equipment 35 years to 25 years ago and hasn't built since. He was just telling me that he just might start building again considering the improvements in parts ie, speed, linearity, etc. I'm sure that you have heard some terrific old amps. After talking with him, I believe that even the old engineers could come up with some terrific products today. I believe a good engineer is a good engineer, we just must wade through so much mediocrity to find them. Good Listening, Tim

"My conclusion is that it is just one of many design and execution factors that affect results depending on how well it is executed."

That seems sensible to me. It's hard to believe that anyone would employ negative feedback in their design if there wasn't, at least in the designer's mind, some net benefit to doing so. Maybe not directly, but perhaps because it allows the employment of something else in the design that more than compensates.

Better yet , how many KT88's would one need for 500 watts necessary for the real thing ?

.......