'In any case choosing speakers based on Sound on a YouTube video alone is very risky business IMHO.'
I'd second that. Throwing an extra transducer between yourself and the recording of the original seems risky at best.
Unless there is a some way to record demos to play back on YouTube which give a close representation of the sound heard in the room. I've seen production values on YouTube (microphones and cameras) look quite impressive, but are they accurate?
It would be great if it was possible and then we could all play them back via a neutral pair of headphones (or monitors) if we wanted to etc.
Perhaps someone can answer that?