Just because one format is flawed does not mean the other is perfect. Yes, analog measures worse than digital. Case is closed in that sense. But that is not enough for those who listen to acoustic music and listen critically.
There are also other possibilities to explain preference, IMO. Just one is that perhaps the digital artifacts are more annoying to certain listeners than are the analog distortions. Fatigue inducing digititis. I don't know what causes it, but I, and many others, hear it on most digital systems. Some engineers have been able to identify what causes this, and have designed gear that is much less annoying to listen to. If the system is sufficiently resolving, these digital artifacts are exposed and heard as such by many listeners. The fact that some listeners don't hear them, or they aren't bothered by them, does not mean that they do not exist.
There are also other issues with digital music that bother listeners. That is why designers continue to find ways to keep improving digital playback. I can't engage in a technical war of arguments because I don't understand the technology the way others do. But I can listen and know that something is not right. So can the engineers at dCS and other companies. The same is true for certain amplifier technologies. One simply must listen to live acoustic music to understand that both analog and digital formats are flawed in different ways. These differences are what lead to particular preferences for individual listeners.
Sure, digital measures well with the techniques commonly used, but it does not sound perfect. It is clear that measurements alone can not describe what we hear. Perhaps one day, they will.
Perhaps putting down the computer screen to read AES articles and sine wave plots and instead getting out to hear a live symphony or local string quartet will illustrate what so many on this thread are expressing. One does not even need to listen to good analog to understand that digital is not perfect.