Ok... busy day but I'll get this answered before heading home.
I went to their web site and did a little research. They leave themselves an "out" by mentioning the following in their documentation.
HOW LOUD AT 8 HERTZ:
While the INFRASUB-18 will reproduce 8 Hz, it is not audible nor does it have enough acoustical power for you to feel it. The measurements are taken at close range with sensitive instruments. To achieve a flat response, full amplifier power is used at the lowest frequencies and very little (<1 watt) in the upper frequency range. The INFRA dual integrator provides the correct signal strength and therefore amplifier power at each frequency. As the frequency is lowered the power and excursion required to maintain a flat acoustical response at high SPL become enormous and not practically attainable for a single driver system. Fortunately the improved phase response, one of the main benefits of an extended subsonic acoustical response, is preserved because the music content is typically not demanding high power subsonic reproduction.
The ideal listening scenario is to have a full bandwidth 8 Hz playback system and play a good recording without low frequency noise present on it. The playback system can then benefit from the improved phase response without requiring excessive power in the lowest octave or engaging the Dynamic Filter circuit.
Not audible is certainly the truth. Their BS about phase is just that... BS. Here is a Eminence Defimax 4018 (18" driver) that will do as well as any other 18" driver for comparison. The 4018 isn't what I'd choose for a sealed box application but it will work fine to illustrate how much driver excursion is needed.
First off LSPCad wouldn't even let me dial in the amount of gain I'd need to get this to work in a three cubic foot enclosure. I'm sure I could get around it if I monkeyed with it but I don't see the point. I moved the box volume up to 67 cubic feet! Yes... just over twenty times the volume of the Bag End. That gives the Bag End a little advantage.
Here we are with 500W @ just 26db extra gain at DC. I'd never dial a Linkwitz transform in with this much gain at DC but to get down this low with a single 18" I had to do the unreasonable. At 8Hz that is equal to about 23db of gain, or about 3800W. You would basically be able to remove the voice coil with a syringe because it WOULD melt if you managed to find an arc welder that could put that much power into it.
Looks pretty good... 115db @ 16Hz. Of course its down to about 101db by 8Hz. This is a 2PI model measured from 1.0M. Looks fairly reasonable if you can live with a 67 cubic foot enclosure.
Here is a shot at the cone excursion to get these levels.
Hmm... little problem. Our cone needs to have about 75mm of X-max one way. If we went with a more reasonable spl level we could get the excursion levels down. I could probably use a Tumult motor on a 18" basket and get 90db @ 8Hz 1M in 2PI space given a large enough box & a huge amp. Of course you would be right at the mechanical limits of the driver and you would be able to get about 80db at the listening position. I'd never design a subwoofer that way. You have no control over what program material is going to play at so you would end up with a smoked driver in no time. If you set a limiter on the system you could dial it in so that you wouldn't clip the amplifier but that wouldn't stop you from going past the mechanical limits. I'm sure they use a sophisticated limiter otherwise they would have a terrible reputation.
This is the reason I laughed at their specs. They are just marketing nonsense. They have nothing to do with a reasonable design and its pretty questionable to cook your specs and report them in a way that doesn't match the industry standards.
