But I do want to highlight the fact that sound is an experience shaped not only by the signal but also by the receiver. In this case a person. I think there may be a significant difference in the perception of sound between a trained sound engineer and an average listener. The engineer (that would be you) is trained to treat fidelity as the Grail. Noise is detected by the trained ear as the defect it is and focused upon. And its presence spoils the experience. But most people can and do enjoy music at wide ranges of fidelity. (And given the right conditions so can a sound engineer I would hazard.) And often they do so under noisy conditions and through noisy devices and using lossy CODECS that leverage psychophysical limits for compression. Witness the average person sitting on a city bus with cheap earbuds digging their MP3 music.
To me this explains the vinyl resurgence over CD. And of course the whole vacuum tube thing. I propose that the nervous systems of untrained listeners may actually like a bit of noise because they are used to it and because the universe is filled with it. For 'warm' translate 'noisy' . Perfect fidelity is desirable in the studio as a starting point. And also by musicians, audiophiles and sound engineers. Was it Neal Young who had his MP3s pulled because he could not stand the sound of them? Personally I am fine with a well made higher bitrate MP3.
It would be interesting to know if there have been any focused tests. I have seen it argued that there will be a generation of people who will prefer MP3 sound over Redbook CD sound. (I happen to think vinyl is noisy in a better way than an MP3 which as we know actually has sound info stripped out of the original.) I will close by pointing out that the average visual system also is also made comfortable by noise. Witness the popularity of the filters on Instagram.