In fact, I'd argue that EQ tends to introduce artifiacts, and smears details and the subtle timing required to accurately recreate a detailed soundstage. Though it may not always seem that way - a cymbal may seem more solidly fixed in the soundstage because you EQ it and bring it's frequencies up.
This is interesting to me, I've been under the impression that you could EQ a suitably flat source to sound however you want. I hadn't thought that a pre could have that sort of influence on the overall image/soundstage on the recording.
Yep, absolutely. The algorithms that filter sound must be implemented in code. Both the algorithm and the specific implementation can have profound impacts on lost detail.
I think those things are noticed more when you record non-PA material that has a complex soundstage. But it also applies to PA material near or on stage.
As far as gain from a preamp, there are bunches of ways to get "gain". Some sound better than others.
I think a few things all need to come together. You need to have a playback system that can image in 3d. It's funny that I put this first, but if your playback isn't capable of accurate reproduction, then you won't know how good your recordings really are, or what impact changes have on the outcome. Also related to that is our age related hearing loss (and how we compensate!), and the overall tone of our playback. If your playback can't resolve low frequencies well, and makes it sound muddy, then you won't want it in your recordings.
A good 2 channel playback will image way outside the speakers. And it will have depth fore-aft. And it might also have vertical components. I've had a great playback for a lonngg time, but it wasn't until I switched to VR4's that the 3d thing became such a big deal. And it also helped that I got rid of my preamp. Eliminating that from the chain had huge benefits - jaw dropper. There is now a large following of folks, and online communities dedicated to the practice of running DA's straight into amps (or with minimal attenuation and impedance matching, etc).
There is a lot of subtle timing and arrival information that must be accurately captured and reproduced to create that 3d soundstage. Any processing or mathmatical operation tends to lose that detail. And that 3d soundstage won't typically reproduce well on headphones because the two channels need to interact in free air.