That sums it up pretty well, along with the additional comments above.
One thing not mentioned is a big difference in PA speaker design, versus either studio monitor reference speakers, or audiophile home stereo speakers. When designing speakers, there are a lot of tradeoffs. One of the biggies is efficiency.
Folks designing accurate speakers don’t worry too much about efficiency, because listening levels are relatively low, and good clean power is available. Quite the opposite for PA speakers. For example, the efficiency of typical small PA speakers is just under 100 dB SPL (measured with a 1-watt signal at one meter). Typical nearfield monitors are closer to just under 90, and sometimes as low as low 80’s.
To put that in context, speakers that are 10 dB less efficient take 10 times the power to be the same loudness. So, instead of that 100W amp being plenty, you now need 1000 W for the same volume. For PA applications, where small PAs run say 500 to 1000W, the efficient speakers are a huge benefit.
That said, I have a number of small PAs similar to what a DJ might use, which I use in bands or for my keyboards (as mains or stage monitors). They make great party blasters. They sound pretty good indoors, played at reasonable levels. But they’re overkill for normal indoor use, and don’t sound quite as good as a good stereo, and not nearly as good as a serious audiophile system (though I don’t count myself an audiophile). They do sound a lot better for that patio party, though, but be sure to invite ALL theneighbors!
I’ve scrutinized that study before, and it looks very well constructed and executed. A shame it hasn’t been repeated by someone else, though, or for people who don’t listen to gammelon music. I’m confident that most of us do NOT hear much over 15K. However, I do agree with those who say live sounds different from recorded.
Not a problem, because you can plug unbalanced outputs into balanced inputs and vice versa. You don’t get the benefits of balanced in either case, and you lose 6dB signal. We do this all the time, though, thanks to having both balanced and unbalanced gear. Less than ideal but no worse than running unbalanced. Note that the balanced inputs & outputs have to be engineered correctly for this to work, but thank goodness they are (or we wouldn’t buy it since we need the compatibility).
I agree with the rest of your post, and take your word for it about amplifier construction details.
A lot of homies use cans these days, and even some pro old-timers (e.g., Craig Anderton), because they can elimiate a lot of variables, and because one can mix while the baby’s sleeping. I don’t recommend it for beginners because you have to understand the huge differences in imaging for speakers versus cans. You’re basically right, except that there are a lot of “nobodies”.