The Impulse test shows up in JA's Measurements from time to time. May 2011's HiFiMan 602 included it and may have been significant: the total lack of pre- and post-ringing may be one reason NOS (non-oversampling, aka zero oversampling) has a certain appeal, despite other horrible metrics (and other horrible sonic attributes).
The Meridian 808 review from April 2009 also included the Impulse response test and associated discussion of apodizing (an minimum phase) filters. In that review, JA noted that part of the reason the 808 sonically excelled may have been control of pre-ringing (ala its software-based apodizing filter). Could control of pre-ringing -- via engineered reconstruction or NOS -- be another 'holy grail' of digital-audio evolution? ... i.e., much like the "discovery" of linearity and jitter in the late 80's and early '90s, respectively? Why isn't the impulse test done more regularly in Stereophile digital-gear measurements?
I'm curious as to how formerly-reviewed NOS gear -- like Zanden and 47 Lab -- perform (methinks it will be similar to the HiFiman 602). Perhaps more importantly, Ayre and Cambridge Audio (both of whom putatively use MP and/or apodizing reconstruction) did not include (TTBOMK) impulse tests. It would be an interesting correlation experiment!