I don't think the value of an SPF meter is that human hearing isn't very good. If it weren't, how would we hear better results by using one?

I think it has most to do with the different speakers sizes and types used in a 5.1 plus setting and most of us not knowing what a multichannel setup should actually sound like.

For example, like most I'm using towers for mains, bookshelfs for surrounds and something larger than bookshelf for center. Though they're all B&W, they sound different when the tone's going through and its harder to compare fine gradiations in loudness to sounds of differing frequencies.

Also, even though I did hear my left main was louder than the right, I never would have dreamed of setting them to different levels (a leftover old wives tail from 2 channel days). And the default of the 950 makes the voice tracks easy to hear by having the center louder, but it turns out to noticeably degrade the overall soundstage unless you set it even to the other speakers (which I think resulted in a center speaker setting of -6 in my system).

I don't think the problem is with human hearing ability, more with our not being familiar enough yet with what a multichannel system is actually supposed to sound like when optimally setup. Which is why the SPF meter works so impressively.

Judgements about how revealing the mids, etched the highs, bloomy the bass, size of the soundstage, has nothing to do with comparing volume levels. And if we weren't capable of discriminating these sounds, non of us would ever need to consider the extra cost and fewer creature comforts of seperates in the first place.