The idea behind the video is good but I have some questions or concerns about he execution and final output. IMO, for a "decent" comparison, you'd want as much consistency as possible during the collection of data. If the sounds were collected from a variety of outside sources, it's very difficult to get accurate comparisons. Even if you recorded all the sounds yourself, what steps were taken to minimize variances? Were the systems on the 2.0 recorded on the same or different cars?
In just the first couple of sounds, the engines aren't run/revved the same and I wonder how many things were the same and how many things were different between them. If it were me, I would try to make sure that every car is parked in the same place, the microphone is placed in the same place height orientation, the engines are run the same, the mileage/break in is the same for each system, time of day, temperature/weather, things like that, etc.
For example, I like how Clavinet Junkie/Oluv's Gadgets does his reviews/comparisons on YouTube. Also, he has this comparable database:
Oluv's interactive speaker audio database