olakunleodebodeWhen I played back just one audio  WAV file of a commercial track extracted from the CD (16 bit - 44Khz - converted to 16 bit 96Khz by SPLAT - and also imported - same upsample into Reaper  - so source audio is 100% identical). With absolutely no plugins - and same audio level - difference was clear.

 
I think what we may have here is an argument based on a false assumption. The source audio may be the same, but the audio files being played back may not be from the same source.
 
What I mean is that if he did it the way he wrote it, two different programs performed the very important task of converting the CD audio from 44 to 96, and one of them might be better at it than the other.
 
All of the counterarguments have been based on using the same source files and comparing the audio playback engines.
 
I've done extensive beta testing on a DAW that I will not name here because NDA, and one of the series of tests that I developed and ran and submitted a bug report on involved imperfect sample rate conversion. It can work just great going up but not down, work great going 44 to 88 but not so much 96 to 44 and so on. I ran these tests on 20-20KHz sweeps and checked the output files using Voxengo SPAN and looking at the resulting waveforms in the DAW itself. There was non-linear frequency response (got louder as it got higher in frequency) and harmonic distortion induced during the resampling.
 
I got the idea after visiting this site, which may be an eye-opener to some: http colon slash slash src dot infinitewave dot ca (I'm writing it out this way because I'm apparently too new to be allowed to play with url's).
 
They get their graphs by simply having a DAW load a sine sweep and convert it to different sample rates and then examine what comes out. The old program known as SONAR acquitted itself pretty grandly, but for some real horrorshows, check out Waveosaur or early versions of Cubase. Gotta wonder how those fireworks translate into audio.
 
To the OP: I buy and rip CD's (and audio files from HDTracks) because I'm a stickler for bit-perfect audio transfer and listening. I use FLAC rather than WAV for the space savings. I've researched the best programs to use to rip and listen to audio from CD's, and I suggest that you are getting nothing from converting a 44.1KHz audio stream to 96KHz except a larger file and the possibility of upsampling errors, which can be audible. And when they are audible, in my experience, they are audible in just the way you describe, where the bad resample sounds flat, less punchy, lifeless, sounds like the transients have the corners rounded off. With it done right, it's like there's more emotional connection to the music, like suddenly the players and singer(s) are in the room with you? It's subtle, but not really subtle, just hard to describe. The best term I've been able to come up with is "blurred transients."
 
My test song is "Everything in its Right Place" by Radiohead. That electric piano, close-mic'd vocal, and little hard-panned sound bites (K-k-k-k-kid-a-kid-a) really reveal quality reproduction.
 
My guess is that if we took olakunleodebode's files and looked at them with SPAN or other tools, there might be a tale to tell.