Loudness is a matter of personal preference.
The goal should be fidelity.
A dozen years ago, when the light bulb went off for me and I realized that sealed + room gain was the only way to achieve accurate reproduction, I built a simple 1 X 12" sealed system and began testing the results.
Luckily, I met Phil Marchand at CES and he showed and explained to me how his Bassis signal shaper worked. I bought one, went home and set about measuring the FR of the sub with no signal shaping and with signal shaping.
Back then, I had to use the RS meter, a CD with 1/12 octave sine wave tones and a C weight correction file and played the tones, noted the reading, added the correction number and plotted them on graph paper I made myself.
I mentioned meeting Marchand was luck because his Bassis is flat to 2 Hz. The preamp I was using was -3dB @ 3 Hz and the amplifier was -3dB @ 5 Hz. I had no idea that was the case back then, so I have to refer to it as dumb luck.
In 2003, I built my 1st dual-opposed sub using 15s and soon afterward I bought a state of the art mic/mic pre/mic PS/Interface that's flat to 2-4 Hz, but I was still limited on the low end as TrueRTA & REW only went to 10 Hz, as did REW. I got Sample Champion by Pure Bits and began slogging through the <10 Hz learning curve. I rarely mentioned that I had the program and had stopped mentioning measurement <10 Hz because of the avalanche of nonsense it evoked. Lots of folks told me the measurements were in error. Wrong windowing, noise floor, HVAC, etc. One respected member actually said a bus or a truck must have been driving by when the measurement was made.
In late '05, early '06 REW extended capability to 2 Hz. Awesome.
By then, Adire had the Tumult. 5+ liters in a 15" driver. I expanded the system to 4 X 15" in 2-dual opposed sealed enclosures, then tried 8 X 15". As Nube mentioned, I posted a formula for building a system to reach reference level in whatever size room in full BW. Using that formula, I was slightly under the displacement requirement with the 4 X 15" Tumult system (just under 22 liters of displacement for 3500 cubes of listening space), but I rarely played back at ref level and have always kept my system calibrated flat with the mains, so it was good enough.
I found that 8 X 15" was overkill for me, but that's where you get back to the 1st sentence in my post. How loud you play is a personal problem that's solved by adding multiples and is a side issue to the question of accurate reproduction.
For reference, it takes almost 2 of the Dayton 18" HOs to equal a single 15" Tumult, which was also equal in displacement to the SI-18" HT driver.
The bulk of my posts since '04 have been focused on <20 Hz playback; is it there, is it intended, how can it be verified as being on a disc, is it audible/perceivable, does it add to the listening experience, etc. Let's face it, <20 Hz is easily accomplished by using the resonant system of one's choice. The 20-20k Hz spec has been around for a very long time and content and playback below that had not been around until relatively very recently.
So, using 8 X 18" HOs may seem grotesquely overkill to most people but that's really only 22-1/2 liters of displacement (rated). Four 15" Tumults, four SI-18s equals that and three UXL-18s equals that, etc.
My opinion has always been that I couldn't care less what most people do or are satisfied with. That has zero relevance and is based entirely on ignorance of the task at hand. Everyone who's exposed to the difference has noticed the difference. It's a simple test, I've conducted the test many times over the years and the results are 100%. This is why, when people say they don't notice any appreciable difference, I simply reply that they haven't experienced the difference.
Here's the result of one of those tests. A 20 Hz ported sub vs a full BW sealed sub. Carefully level matched where they are flat. Same scene played A/B in unknown order and asking the listeners to describe the difference, if any. This is what it looked like from the in-room FR and mic'd SL waterfall perspectives:picasion gif maker
In my room, using very accurate measurement hardware, the ported subs do not have usable output below tune. This differs quite a bit from what many ported sub owners claim, but the data is what it is and has been consistent with 4 different well known ported commercial subs over the years.
GTGs usually become loudness wars. The problem with that is there is no data gathered other than how loud each sub played. Commercial subs use sophisticated limiters that radically change the FR of the sub, so as level is pushed up, the low end begins to disappear. A simple SpecLab cap would show this phenomenon, but so far, no one has semed willing to set up for and gather that data, leaving only subjective impressions, for what that's worth.
The requirement is to have a signal chain that feeds the content to the drivers, to properly set the voltage gain of the components in the signal chain, to have enough displacement for the drivers to respond within their operating parameters, to properly shape the signal to modify the native response of the subs to properly mate with whatever room gain profile and to have enough amplifier power to amplify that boosted signal within the amplifiers rated operating parameters.
And, yes Steve, I believe that ref level in a bedroom is just silly. I use a single 12" system and it's plenty.