I don’t mean to pick on the OP or engage in Audio Heresy, but it’s interesting that he has been listening to a compromised source, namely Spotify Free, and enjoying it. It isn’t until others have been pointing out how compromised his source is that he is starting to become seriously concerned. Now, the OP states he is a veteran listener, and speaking as someone who just flunked his hearing test, some of that may be due to an age related loss in auditory acuity. However it is more likely that Spotify is doing its job with its algorithm.
I have spent a large chunk of my background listening the last 10 years to a radio station, Radio Venice, that sounds pretty good. Fire up a CD on the same system and the improvement becomes obvious, but listening in isolation, I can and have whiled away time listening with much pleasure. I was surprised to discover, when I purchased a streamer that displayed bit rates, that it broadcasts at 128 bps. Listening critically I can tell that my local Classical station, WFMT, which broadcasts at 320 bps, sounds better, in terms of more air around the musicians. However, instrumental tone is largely preserved at the lower bit rate. I always thought that iTunes over a mid Fi system or headphones sounded pretty good. There are some low bit rate stations however that sound awful. The BBC Radio3 probably sounded better to the French Resistance listening clandestinely in WW II than I can get. I don’t think that bit rates tell the entire story. There have to be differences in the algorithms that make a significant determinant in audio quality. I used to subscribe to Spotify years ago and always thought they sounded pretty good, and remember thinking that Ogg Vorbis must be a superior compression system.
My main experience with upsampling comes from a DAC that I used about 15 years ago, when I know my hearing was better. It would upsample to 96 and 192 I listened mainly to CDs sourced from whatever my CD player was back then. I remember for the most part being impressed by the lower upsampling setting. It seemed to expand the soundstage, particularly in the front to back positioning of the Orchestra. The higher upsampling rate however sounded off from the get go. It seemed like the same soundstage was moth eaten, as if there were gaps in it. Eventually I just went fo no upsampling.