I’ve never really understand the usefulness of blind tests. Do people actually know what they prefer in a short listening A-B session?
I agree that this is potentially problematic, but to be fair, such tests are not about preferences, but differences. In other words, can the listener simply and consistently distinguish between components, irrespective of which they may prefer.
But while the science suggests that humans have very short auditory memory, I'm with you on the value of listening to longer stretches of music in order to make confident judgments about differences in components.