Are our 'test' records adequate?

Most of us have some favourite records with which to check the health of our systems, or to assess a new component within our systems.
These records are often carried with us whenever we wish to assess a completely foreign system in a different environment. I have my favourite ‘test’ records, some of which I continue to use even after 30 years. I know them (or parts of them), so intimately that I feel confident in my ability to assess a component or complete system after just one listen.
I know other audiophiles who have specialised their ‘test’ records to such an extent that they have different discs to evaluate for Voice, Bass, Large Orchestral, Chamber, Piano, Strings, Drums, Jazz, Rock.
Almost invariably, these vinyl discs are superbly recorded and sound stunning, not just on very fine systems, but also on average systems.
Of course, because each of us knows his own discs so intimately, it is possible to assess the 'omissions'in a foreign system by memory, often to the puzzlement of those to whom the discs are not so well known and to whom the sound had been thoroughly satisfying and impressive?
But I have begun to wonder recently, if this is in fact the most reliable method of evaluating components and systems?
I am sure most of us have heard records on our systems which are almost unlistenable or certainly unpleasant and we have simply placed these discs in the 'never to played' shelf of our storage unit?
But perhaps some of these records might be more revealing than our fabulously recorded 'test' material?
For some time I have been disturbed by two records in my collection, which despite their fame, have sounded poorly (in various parts) despite improvements to my turntable, speakers, amplifiers and cartridges.

Harvest by Neil Young on Reprise (7599-27239-1) has some nicely recorded tracks (Out On The Weekend, Harvest, Heart Of Gold) as well as 2 tracks (Alabama, Words), which have confounded me with their leaness, lack of real bass, vocal distortion and complete lack of depth. The album was recorded at four different venues with three different Producers and those two tracks share the same Producers and venues.
After mounting a Continuum Copperhead arm as well as a DaVinci 12" Grandezza on my Raven AC-3 and carefully setting arm/cartridge geometries with the supplied Wally Tractor and Feikert disc protractor, I was actually able to listen to these tracks without flinching, and could now clearly ascertain the 'out-of-key' harmonies of Stephen Stills together with the clearly over-dubbed lead guitar boosted above the general sound level on the right channel and the completely flat soundstage.

Respighi Pines of Rome (Reiner on the Classic Records re-issue of the RCA LSC-2436) had always brought my wife storming down the hallway at the 'screeching' Finale whilst I scrambled for the volume control to save my bleeding ears.
Again with the two stellar arms and strict geometry, the 117 musicians could not hide the shrill, thin and overloaded recording levels of the horns (particularly the trumpets).
But the wife stayed away and my volume level remained unchanged.

My wonderfully recorded 'test' records had sounded just fine with my previous Hadcock arm but it's only now, when two 'horror' discs can be appreciated, that I truly believe my system 'sings'.
Perhaps we could re-listen to some 'horror' discs in our collection and, with some adjustments to our set-up, make them, if not enjoyable, at least listenable?