I can take the subs out of my blazer and stick them in a tiny trunk car and they will certainly get quieter. Also, to back me up, when you roll your windows up, you are letting the "pressure" escape and now have infinite space instead of just a tiny car. The bass still comes. This is because the hatch is acting as a parabolic antenna. I can use logic all day long to disprove you.
That's not really logic. Trunk cars are hard to get loud because the trunk physically holds the sound in the trunk keeping it away from the cabin. This has an even bigger effect than cabin gain. However an open small vehicle aka a hatchback is usually louder than an open large vehicle, aka an suv, try holding variables somewhat constant if you using logic. Heck cavaliers are usually known as one of the louder trunk cars, they also dont' have a great seal between the cabin and the trunk and are relatively tiny for trunk cars......
When windows go down your effectively creating a ported box inside the car. The windows act as a helmholtz resonator and amplify more audible frequencies. If you actually look at overall energy though, you do lose by opening a door or window. It's all 20hz and down though. John Janowitz from AE speakers did measurements on several cars and found this to be the case every time. If you want more proof look at the extreme cars like alan dante. They are super tiny on the inside and completely sealed. They take advantage of the cars resonance frequencies and then seal up and pressurize a very tiny area to hit silly high numbers, often with very little cone area (world record right now is 2 18's for **** sake)
The part your wrong about is the wave development. A soundwave doesn't have to fully develop within a space in order to be loud. If it did no car would have good SPL levels at 20hz, after all a 20hz wave is over 50 feet long. Yet every car has extreme cabin gain near 20hz, usually upwards of 30 decibels or more. It's because the wave cannot develop in inside the car without reflecting back onto itself multiple times, increasing it's overall pressure. Again the basic formula is (speed of sound)/(frequency) = wavelength. Given most vehicles are 10-15 feet long that means a wave would fold back on itself 4-5 times give or take a bit. Each doubling on itself is 6 decibels of output. 6x5=30, which again looks pretty close to basic theory estimates. Now if the wave has to fully develop then cabin gain should be crap at 20hz and we should be using super small sealed boxes in the home in large rooms as according to you, they would be able to support a 55ft wave, meaning it would be louder and big boxes tuned low wouldnt' be needed.
Anyway I've shown you links explained the mechanism and even showed that the well known formula used in both home and car audio holds constant with the real world. If you would like to believe pressurizing a small area is harder than a bigger one, go ahead....
Lastly any sort of parabolic antannae doesn't apply here. The relatively short wavelength of electromagnetic radiation at these frequencies allows reflectors to have a highly directional response for both receiving and transmitting. A soundwave in the subbass frequency can't be aimed like that, not without a VERY big opening. Go look at a true subbass horn, the horn mouth is massive even to get directional response down to 40hz. Again a soundwave that long doesn't really see the enclosure it is in, the waves are too long. It's simply pressurizing a room smaller than inself folding back on itself at it goes, increasing the pressure. Not to say you can't add SPL by aiming a sub in a vehicle but no matter which way you aim a sub in a hatchback the cabin gain will be better than in a much larger open spaced vehicle.