The Best Preamp is no Preamp?

That means the pre does change the original music signal into slightly different signal. The modified signal would be good to our ears. Do not praise the preamp that makes poor source better. Just blame the poor source. If you have good source (well-recorded CD for example), you can enjoy the music much better with no preamp. Why does tube-based gears make good sound? Just because the tubes tend to add some distortion to the original signal, and the added signal makes our ears pleasing. The added signal is usually even-order frequencies.

Shorter speaker cables make the over-all sound better? I have to laugh at the guy who says that. Inside the speakers, there are very long coils. Due to limited space condition, the coil's physical shape can not be better than that of speaker cables. Compromised long speaker voice coil and very short speaker cable? Usually, the speaker cable length does not matter very much unless the cables run longer than 50 feet.

A preamp may well not load a source as heavily so the source may sound better due to wider bandwidth and lower distortion.

The choke in a speaker is something very different from the speaker cables! Speaker cables have low inductance while a choke has high inductance. But if we are to take this argument on its face, the correct conclusion would be that since a choke as able to cause highs to roll off and its substantially similar to a speaker cable, the conclusion would be that you really don't want long speaker cables! 

In practice, long speaker cables often cause a loss of bass impact and intelligibility of vocals. This is easily measured and demonstrated. 

but is there any explanation for how putting a preamp in the signal path when you have a fairly higher-end DAC can make an improvement?

Yes. The explanation is offered earlier on this thread.

This has been an on-going topic for a very long time! @phantom_av makes a good point about the quality of the line stage; IME if a passive sounds better its telling you something about the line stage you used for comparison. IMO line stages have been in a sad state of affairs for a long time but if you get a good one there's no going back.

I don't like the problem were a passive sounds fine at full volume (which becomes the source driving the amp) but as you turn down the volume even a little bit the bass loses impact was well as the dynamic character overall.

If you never are able to turn the volume up all the way with a passive you may never find out what you are missing.

You can avoid this problem to a limited degree by using a control of a lower value for example 10K instead of 100K. But at that point a lot of sources will choke as they are not meant to drive impedances that low. 

In addition you have to keep your cables short and be picky about what cables you use. I found out decades ago that if you can place your amps by the speakers and run short speaker cables you get an instant improvement in resolution across the board- but only if you can get the signal to the amps intact. I ran balanced lines for that and so had no problems. So for the last 30 years that is what I've done, and no worries about what cable I've used. They are 30 feet long!

That simply isn't something you can do if you use a passive control. If you have a smaller situation where a meter cable will be long enough then if you're careful (or lucky) you can get it to work quite well. My bedroom system employs a passive control built into the power amplifier which IME is the best place to put it.

Thus, the only explanation for this is that the brain “likes” certain types and amounts of distortion. A more complex topic is how distortion from different parts of the chain interact.

@tangramca There is more than one explanation! If you've ever heard differences between interconnect cables then you know that they can color the sound too. A good linestage can vastly reduce this coloration (on account of its lower output impedance). If the linestage is doing its job properly you might not hear any significant difference between interconnect cables at all.

Sound reinforcement systems don't have pre-speaker-amp "preamps" as a concept even.

Of course they do! The preamp is known as a 'mixer'.

*******************************

There are four functions a preamp provides:

1) provide input switching

2) provide volume control

3) provide any needed gain (tuners usually do not make enough output to drive a power amp directly; many make only 1 volt)

4) (and the least understood) control interconnect cables; prevent or reduce them from adding colorations due to cable construction (if you've ever auditioned cables and heard differences you know what I'm talking about).

If you use the balanced line system, you can reduce cable artifacts and should eliminate ground loops entirely. The balanced line system is intended to be low impedance which rules out passive systems. I tried this decades ago and found dramatic differences between cables that were eliminated when I used an active line stage that drove balanced cables properly (e.i. supports AES48, the balanced line standard).

If you run short cables you can get away with this. But if you have monoblocks (which are useful for keeping your speaker cables as short as possible) then no so well. I've found that using balanced connections to the monoblocks and having short speaker cables got me better bass, more engaging midrange and easier to understand vocals.

You can't do that with a passive- it allows the cables to color the sound. Quite often sources like a DAC can't handle long cables either. But I've found over time that the quality of the preamp has an enormous affect- and the length of the cables has nothing to do with it. Some are terrible- and so using no preamp at all is better, and some are great and actually sound a lot better when used.