> high SPL the room should not matter...
An "obvious" point that isn't obvious to everybody (even if they done a lot of sound):
The room-sound SPL is proportional to the direct and near-field sound. Play louder, the room sound is louder. Logical, since the room sound IS the direct sound after many bounces inside the room.
There is compression in a tube guitar-amp at high power: short-term compression from grid overdrive and longer term compression as the voice coil gets overheated. I don't see how this changes the direct to reverb ratio. There is another kind of compression as you begin to approach 194dB SPL, but the cone of a super-power speaker isn't much over 135dB SPL so that does not apply. (194dB SPL only happens in the near-field of a bomb or sonic-boom, and you don't want to play there.... the throat of a horn can hit 150dB SPL and compression causes distortion in the throat even if the driver is perfect, but we don't mike horn-throats for music recording.)
The fact that the room-sound is the "same all over" isn't "obvious", but very good to know, and even to prove to yourself.
Get a small sound source, like a computer speaker, micro-monitor, or one of those $29 9-volt "guitar amps". If it has weak bass, that's good: most rooms get erratic below a few hundred Hz so we'd like to avoid exciting bass-modes for experimental clarity. Find a sound that is constant-volume: pink noise (or FM noise) is traditional, but a heavily-compressed pop track works fine. Something where a VU meter hardly moves. Then get a sound-level meter or a mike into anything with a good meter.
The "same all over" effect is easiest to see in a "hard" room, a concrete garage or cellar. You can observe it in any room live enough to sing well in, and should try your music room, but a hard room is easier to get the idea.
Set the speaker somewhat off-center in the room. Put the mike very-near, and adjust the speaker for pretty loud (about as loud as you or the speaker can stand, for best S/N), and adjust the mike-meter for a reading near the top of the scale. Be sure the source-sound is constant-volume enough so the needle doesn't waggle much. (On a sound level meter, use the Slow setting, and A-weight emphasizes the midrange where we hear best.)
Now go all around the room, watching the meter. The level almost everywhere in the room is within a dB or two of any other place, and about 15-25dB down from the near-field level. There will be some rise near walls and corners, especially in the bass. There is rise as you approach the source, of course. But in a hard room, ignoring bass and the area near the source, the level all over is pretty darn constant.
Why? Look at the sound. It comes off the speaker, spreads, and weakens with spread. 6dB every time the distance doubles. But then it hits a wall and comes back the other way, still weakening. And then it hits the opposite wall and comes back again, overlapping. If you sketch it with "pressure" on a vertical scale and the left-right layout of the room, you have a lot of slanting lines, half slant down-left, half slant down-right. If you figure the power-sum of these slanted lines, you get a level line: same pressure everywhere. And a live room will bounce sound for about 0.3 seconds before it decays 20dB. If the room is 20 feet, there are about 15 bounces before the decay is 20dB. Stand in one spot and hear the same thing 15 times, weaker each time. And more reflections from the other walls and the ceiling/floor, 50 audible end-end bounces in a typical "nice" room. So it really IS the "same all over".
Now walk from a far point toward the speaker. Note where the level seems to rise 3dB. That distance is called "critical distance" by some acousticians. Keep going, and there is a range where every time you are half as far, you are 6dB louder, simple direct-sound inverse-square relation. For large speakers, when you get within about one cone-diameter, the level stops rising: 1" is about the same as 2", and for a 12" speaker a mike at 5" or 10" from the center may not show much change of level.
That "critical distance" or Dc is where the direct and reverberant sound levels are equal. Going closer is louder, going further is all the same level. The direct sound continues to drop: at 3 times Dc the direct sound is 10dB below the room-sound.
This is "critical" in speech PA systems: listeners should not be closer than Dc to avoid blasting, nor further than Dc because room-sound starts to smother speech clarity. Also the potential acoustic gain from talker to listener depends on Dc (and some other things).
For true stereo recording os an acoustic act, like orchestra or choir, where the room is part of the act, this same range of Dc to 3*Dc is a good place to set the mikes. Closer than Dc will "close-up" the band and often be too dry. To a live listener, music can be good well beyond 3*Dc because most music does not demand the spit-clarity that communicating speech needs. But for record-playback where the listener does not have sight and head-turn clues to the acoustic, 3*Dc is normally rather far, 2*Dc is about as far as I'd go unless the room is very nice and the music can be "room-lush". (I guess that organ gig posted in the Pub was done about 8*Dc, but organs evolved for such spaces.)
Acoustic theory assumes the source and mike are omnidirectional. A very directional source like a Fender Twin will show a larger Dc in front and shorter Dc on the side than a true omni source, and directional mikes also distort effective Dc.
Dc is short in a live room, long in a dead room. Outdoors, Dc is infinite (no room reverb at all).
Dc in a typical US livingroom is 3 or 4 feet. Many home-studios will be similar, unless they over-do the wall-foam which will increase Dc. Concert halls 150-800 seats tend to cluster around Dc of 11 feet, and Dc does not rise much with size because if Dc were too large, levels in most of the seats would be low. Large concert halls have a lower percent absorption than small halls to keep room-sound levels high. Pro studios will range between livingrooms and concert halls, with wide variation at similar size depending on flavor "live" or "dead". Whether the main use is musicians who want reverb for blend and self-feedback, or players who need to see each other without hearing each other or cross-talk into each other's channels and reverb is added later.
Oh, and for the public performance situation, there is a fairly simple way to estimate Dc at a glance. Most for-money playing happens in rooms where the walls and ceiling are hard (plaster, mirrors, hard paneling, beer-signs) and the floor is "soft" because it is "covered with people" (can't get rich in a big room with a small crowd). In a concert hall with about 75% of floor covered in soft bodies (auduences have high and pretty predictable absorption) the Dc is about 1/4 of the ceiling height. Clubs can be a little less with tables, or a little more when the joint is packed butt-to-belly. (A crushing-crowd absorbs less than an elbow-room crowd, but a crush-crowd is a disaster waiting to happen. Someone here has a link to a soccer-disaster.) You see that in a long low club, much of the crowd is beyond 3*Dc and you need a lot of "throw" (directionality) in speakers to punch clarity to the far corners. (But maybe those folks picked the corners because they don't want to hear the band so well.)
I've digressed from your close-mike cabinet to smokey clubs. But the sound waves don't know the difference, and the general relation between direct sound and room reverberant sound follows the same general trends anywhere. Put your ear to the jukebox and you still hear the murmur of the crowd in the whole room. In that case turning-up the jukebox should overwhelm the crowd (except real crowds talk louder when the jukebox is turned up). When the room sound is the decaying direct sound, turning-up or -down won't change the ratio of direct to room sound (unless you turn down so very far that background noise downs the room reverb).