I consider the whole 90% thing to be a relatively meaningless abstract, explainable but not quantifiable. I mean, 90% of WHAT, exactly? The capabilities of any particular system, in a particular room, can be measured. All individual components of a system have specs that can be measured. Performance has an absolute 90% cumulative value score in any system, it's a real set of numbers, and I think 95% or so is a realistic performance goal. But, just because a component is operating within spec DOESN'T necessarily mean it's operating at maximum capacity or efficiency at any given time, and doesn't take into account the rest of the components in the system. But how do you measure the enjoyment of music? Your system may measure off the charts good, with everything working at or near its maximum capabilities. But if I don't care for your choice of speaker, say you have horns and flea-watt tube amps, and I prefer planars with big honkin' SS amps, you might only be at 90% (85% if I'm really being snooty!) in my estimation. Which of us is right? Also, if you change any component, from cables to speakers, you now have a different system, with slightly different performance parameters, whether you can hear them or not. Do you take inaudible improvements (.01% vs .02% THD) into consideration, or is it "only improvements you can HEAR?" And how do your experience and expectations color your opinion? The goal line is always moving. Think of how boring it would be if everybody liked exactly the same things, and thought 80% was good enough. We'd all ride the bus to work, eat at McDonalds for lunch, pizza for dinner, and listen to canned music on a phone with some phat Beats.
Bottom line, quit worrying about numbers. They're not really rules, so much, as guidelines...