Being a bandpass I anticipate it having a giant peak like response and for that reason I have always dismissed it thinking it will be a one note wonder.
Bandpass designs can sound as good (or bad!) as any other. I built one a little while ago (the chosen drivers practically forced a 4th order bandpass), and it sounded great once properly integrated.
A "gotchya" is that the bandpass forms its own acoustic crossover slope at the top end, so you've got to factor that in when programming your processor. A 12dB/octave Butterworth (electrical) slope might get you a resulting 24dB/octave Linkwitz-Riley.
There are plenty of badly-designed bandpass subs, and plenty of well-designed ones being used badly.
I quite liked the 4th order bandpass for the following advantages:
- No particular need for an infrasonic filter - it would provide some output at 10Hz on account of the sealed back chamber.
- With a pair of drivers, you can set the internals up for force cancellation. Bonus points if you get the port(s) passing cool air directly over the magnets.
- Smaller than a well-aligned ported box
- Lots and lots of output, albeit not going as low as a bigger ported box.
A double-edged-sword is that the acoustic bandpass filters out most of the distortion from the drivers. The sound stays pretty clean until the drivers are seriously struggling (and likely about to fail).
Chris