I definitely respect your opinion Peng but, I don't agree with it in this case.
I think it totally depends on the Mic or SPL Meter he is using. The Audyssey mics are cheap, medium quality at best, and batch calibrated within a variance of 2.0db. If he is using a UMIK-1(especially if calibrated by CSA) or even a quality name brand SPL, that is accurate at the right frequency range to measure a sub, they both should be more accurate then the cheap included Audyssey mic(some peoples SPL Meters are worth half as much as their AVR's).
I am new to this and still learning but, I have read posts from several people that say you should indeed re-check levels post Audyssey and correct where necessary. Relying on a single measurement, with a cheap mic, during the Audyssey position 1 scan is not the most accurate way to do things imo. All it takes is a car driving by or any other issues and that one scan will not be fully accurate. With a quality SPL Meter/UMIK-1 & REW you can play a constant tone and recheck it until you get the level perfect. Then, you can recheck it anytime. While Audyssey does a pretty good job, I can't see how this method is not more accurate. I personally see Audyssey as a basic starting point(or finishing point for less technically inclined people)that can almost always be improved apon.
As for his question about the subs...
I believe you want the combined output of all three subs to be 75. But, that is just a starting point. Adjust the bass, post Audyssey, to the level you see fit(many find it too low after Audyssey). Just make sure to keep your AVR sub levels at about -6.0db or lower to prevent the AVR from sending a clipped signal. Use your subs gain to go higher then that while keeping them all synced.
Sent from my SM-N975W using Tapatalk