What limit do you use for your "3 dB under the limit" rule? Xmax (or something like CEA score) or Xmech?
For pro use, mostly > 40 Hz, these IPALs are undoubtedly awesome. For ULF sealed sub use, they are very good for getting a lot of output in a small space, but the low minimum impedance holds them back, IMO, compared to other options. The differences are very close though. If SP made a 10 kW X 2 channel amp, it would probably work very nicely on 6 of these, wired as triples in series for a minimum impedance of 2.1 ohm.
The resonance could be in the surround too. Glad to know you don't hear it, so it's likely mostly harmless in this case.
All of the above...That's what all of the data for raw drivers here is about is providing useful information on how the drivers behave and where they start to fall apart and using it to make informed decisions on how they can best be utilized in a finished design. What I meant above is I prefer to operate at 50% of capability or under if at all possible. I like to WAY overshoot the headroom requirements in the drivers used whenever possible. The moving coil driver is usually the weakest link in the chain other than the acoustic space. I don't want to plan on using the last 3 to 5dB headroom of any system. In that way, when looking at the measurements here, I use them to judge where that point is and what the results should be in the proposed design.
I see people all the time who advise against using a higher performance driver because they don't have enough power on tap to "push it" because they only have a 1500w amp and winisd or whatever says they need 3000w to go past xmax a bit. Then someone will advocate using a cheaper driver because it will "match up" better with the amp and will reach xmax on the 1500w amp in the simulator. They often forget the a 6dB increase requires double the displacement from the driver and a 3dB increase requires 2X the amplifier power. That last 3 to 6dB has the reproduction distorted the most by far. The less capable driver may be operating safely but far less effectively near the edge of its performance envelope at times, while the other more capable driver is loafing along. There's nothing wrong with having enough amp to overpower the drivers by any means of course either as I'm definitely guilty there, but I don't like to be operating at or even near the limit if I can help it.
If I could get B&C to make changes to this driver it would be #1 to make it available in either 2 or 4ohm if possible. #2 would be to soften the spiders a little. The Fs up near 40Hz is not really an issue but if it was allowed to be in the 25Hz range that would help get the impedance peak down lower in frequency. The Qts and Qes would drop more as well. Obviously these are about as far from a typical sealed driver as you can get and these are not intended for that in any way shape or form. They do well simply because the xmax is generous as is the power handling and the motor is ridiculous. However with the impedance peak up near 50-60Hz in a sealed cab the amp better be able to dump some current below 30Hz. Just not a sealed driver really. However it can work and is quite fun. Clearly optimized for 30-100Hz range. Now once we get these into vented, BP and horn cabs things get interesting. Due to the motor force the response shapes come out rougher than normal. You have to understand that what you are seeing as peaks in the response is increased efficiency and that these are made to be used with modern signal shaping. For example the peak at vent tuning you'll see in a typical cab size that could support the port and tuning needed for one of these, looks bad on the raw unprocessed graph. However if you add in the HPF and LPF that will be needed at minimum things look much different.
Bit of trivia...This driver was developed years ago and is the grandfather of the current line of B&C woofers. The 21sw152-4, SW115's, SW150 and their other high power neo subs were developed after years of work on this one and using the knowledge gained from it. I had thought it was a buildup to producing this design but it was the precursor.