OK, well some test equipment can, and some can't. Whether the test
equipment
exists today that is good enough to tell what I can tell with my ear is
one
question (I don't know whether it exists or not.) Whether that is the
test
equipment that is actually used to do a review is another thing. (I'm
sure some
reviews have been done with cheaper, less accurate equipment.)

It depends what you refer to when you say "what I can tell with my ear". If
you're referring to detection of harmonic distortion, detection of noise, or
detection of IMD, then I can tell you with certainty that we've long had the
test equipment to be able to perform such simple measures. In fact, all you
need is a high quality microphone to capture the signal, and the analysis
can be done on any PC (with free software, no less!). But if instead you
mean whether or not we have test equipment to detect the mood of someone's
voice, to detect a voice as someone you know under all conditions (eg. even
when their voice is altered because they have a cold, etc), or to translate
a voice into speech, then no, the human brain is still superior to test
equipment. But again, I'm not talking about determining which sound is
better, more musical, warmer, brighter, smoother, realistic, etc. I'm
simply referring to whether or not one can detect a difference under a
controlled setting.

As for the equipment that's used to bench amps, nothing terribly expensive
or elaborate is required. In recent years, I've used expensive data
acquisition cards to do that sort of thing. I've used simpler USB data
acquisition devices. And, even more recently, I've used a good sound card's
input, a homemade attenuator, and some custom written matlab software to do
it. People who benchtest for a living undoubtedly use more sophistocated
equipment, but it's not necessary IMO. In short, you can get incredibly
precise equipment for extremely low prices. Hell, Arny Krueger benches
soundcards WITH a soundcard, and his tests are among the most comprehensive
available!