is McIntosh known for good dynamics?

While I understand what you're saying, that distortion will make the contrast between loud and soft greater, I don't think that is likely to result in realistic dynamics.

I have listened to some good SET  headphone amps, and they had great microdynamic resolution.

@magon  Your comments here describe exactly what I'm talking about, in particular your comment about SETs. This is literally one of the things I have against them! (The other thing about SETs (not so much headphone amps) is they have real problems playing bass, but that's a discussion for another thread.)

Now to understand how this works you have to understand how the ear differentiates sound pressure levels. It does this through the detection of higher ordered harmonics.

This is very easy to demonstrate using simple test equipment: 

sine/squarewave generator, 

amplifier

a speaker and a VU meter. 

You start by connecting the VU meter in such a way (perhaps at the output of the amp) that it can read easily across its entire scale. Then  you put a sine wave through the setup, set the gain to read zero VU on the meter (near the top of its scale). Get a good idea of how loud that sounds to you. Then you turn the volume all the way down, cover up the VU meter, switch the generator to square wave and turn up the volume until it sounds as loud as before. Then uncover the VU meter and you'll see what this is all about.

Square waves are composed entirely of higher ordered harmonics.

If you don't understand that the ear uses higher ordered harmonics to sense sound pressure, then what I've been telling you would make no sense.

so with no other explanation available to the engineer, they say that it's "euphonic distortion."

Actually the explanation has been available to the engineer for nearly 100 years (I refer you to the Radiotron Designer's Handbook, 3rd edition, published in the 1930s). Here it is: 'Euphonic distortion' is lower ordered harmonics, the 2nd and 3rd.  They are innocuous to the ear in that they are close to the fundamental tone and musically related (the octave above and the 5th above that). So they are unobjectionable and simply contribute to that quality audiophiles refer to as 'warmth' or 'bloom'.

@atmasphere 

You don't seem to have gotten my point. TLDR; distortion can't explain realism.

Longer explanation: what I call good dynamics are also realistic dynamics when comparing the sound with live classical music, which has the greatest micro and macro dynamics of all. 

Second point: any reproduction system (including photographs, video, painting, whatever) has distortion and also can be realistic to a varying degree at the same time. We're pretty good at ignoring certain kinds of distortion. For example second harmonic distortion might be largely ignorable or have no impact on the musical presentation. Any system can also be accurate in some respects at the same time it has distortion. For instance, system A might have a non-flat frequency response and system B might be flat as a ruler, but system A might have more realistic transient reproduction. 

I find good vinyl to be more realistic than most digital. The engineer might say "vinyl has more distortion, therefore you actually are liking the distortion." But I didn't say I like vinyl, I said it's more realistic. The engineer has no explanation for that. Therefore they fall back on the euphonic distortion theory. I don't think the engineer is interested in finding out what vinyl does more realistically. Maybe its transient reproduction is more realistic. 

I found the SET amps to have realistic microdynamics. My belief is that to the extent these SETs had distortion, it's largely ignorable or doesn't affect the music. By the way I'm not making a claim about all SETs as I'm sure there are some with higher distortion, bad design, or plain don't sound good. 

what I call good dynamics are also realistic dynamics when comparing the sound with live classical music, which has the greatest micro and macro dynamics of all. 

@magon 

I’ve no problems with that but I would not use the word ’dynamic’ in its plural version, to avoid its common use when distortion is really what is being unknowingly discussed. 

If you had done the test I outlined then you would not have responded in the way you did. Your belief about SETs is just that. When you take away the distortion SETs make you take away their ’dynamics’. In a typical SET running no feedback, you really only have about 20-25% usable power before higher ordered harmonics are showing up on the leading edges of transients (where the power demands are). Most people use speakers that are far too inefficient to avoid this problem! For example if you have a 7 or 8 Watt SET (likely using a 300b) the speaker needs to be around 103dB at 1 Watt to really show off that amp. That’s a big horn speaker and most people don’t want them taking up the space. So you read online how such amps are so ’dynamic’. Keep in mind that such amps also make typically 10% distortion at full power.

If you get the opportunity to hear a PP tube amp of the same full power rating as the SET (IOW, leveling the playing field a bit by reducing variables) you find out really quickly that SETs have nothing over a properly designed PP amp. Usually when SETs and PP are compared, the class of operation (class A) is ignored along with the power tubes used, use of feedback and the total output power. If you eliminate these variables the advantage of SETs goes away completely using any metric that audiophiles use. But finding a PP amp that is low power, or class A, or uses the same power tubes all at once is nearly impossible.

So we have comparisons of 300bs or 2A3s running class A against KT88s that are running class AB with feedback.

In case its not clear I’ve been designing tube amps for a living for about 50 years.

I find good vinyl to be more realistic than most digital. The engineer might say "vinyl has more distortion, therefore you actually are liking the distortion." But I didn’t say I like vinyl, I said it’s more realistic. The engineer has no explanation for that.

This engineer does...

I ran a mastering operation for about 20 years. My setup (Scully lathe, Westerex 3d cutterhead with 1700 series electronics) employed a 30dB feedback loop around the cutterhead and mastering amplifiers. When you run that much feedback, the actual distortion is quite low and is certainly much lower than the digital community would have you believe. 

The simple fact is that most of the distortion comes in during playback, starting with a poor choice of platter pad (if the LP and platter are resonating all bets are off), problems with setup, problems the arm design and problems with the phono preamp. They are all solvable. The more they are solved, the more relaxed and musical the vinyl becomes.

For several years I’ve been using Focal Sopra 2 Speakers driven with a Simaudio 700i integrated amp.  Last summer I upgraded to a Macintosh MA 9500 integrated amp using Cardis Clear Reflection Cables with an Esoteric K-03 XD SACD player. While the 700i has a lovely sweet and detailed sound when I switched to the MA 9500 the change in dynamics was instantly noticeable and much better. There simply was no comparison. I heard deep bass that I never heard before and the warmer Macintosh “House” Sound was far more suited for my Sopra 2 Speakers. 
I auditioned a Pass Labs 250 at the Dealer and I preferred the Macintosh with the Sopra 2.  The Pass Labs sounded similar to the Simaudio, which was not bad, actually very nice but the Focal/Macintosh combo was a match made in Audio Heaven.  The Focal Sopra 2 is a very efficient 98db speaker and the sheer power and dynamics of the MA 9500 made them come “Alive” in ways I never imagined. 
They sound more like the much larger Sopra 3 speakers.