A post I recently made to another forum might possibly be of interest here (the other forum members just yawned and said "Not interested, let's move on...")
The problem with spending a week or two listening to new kit or layout then swapping back to the old, is that over this kind of period our perception habituates (in the physiological, psychoaccoustic sense), i.e. we become accustomed to the new sound. The effect is much as if you turn the bass boost up high and listen for an hour or two - it no longer sounds so loud, and when you reset it to normal levels, it sounds thin and weedy. Habituation effects occur in pretty much all our senses, in differing ways and over differing timescales. I'm pretty sure this is the expanation behind many of the 'burn-in required' claims. As I understand it, physics/electronics suggests that only mechanical or valve-based components burn-in over periods greater than a few milliseconds, and for mechanicals, the burn-in period should be short and should be achieved during factory testing or very shortly thereafter.
It seems to me that if we don't have some reasonably well-trusted method of comparing different components or layouts over short periods (surely this is what techniques like tune-dem are for?), we find ourselves in a catch-22 situation. If we listen to the new setup for several days, we will become habituated to its sound. If we switch back to the old system, it will sound different until we have habituated to the current one. This means we can't make an unbiased decision based on what we percieve immediately after switching systems if we don't have a means of assessment that is reasonably independent of these effects. Further, if we leave it for another few days, we are in the situation where several days have elapsed since we heard the other system fully habituated. Even if we have rated each system on some scale and kept a written record, there is no guarantee that our physiological and mental state is comparable over each rating period, even assuming the room itself is unchanged (moving furniture, books, or stacks of magazines, etc., can noticeably affect perceived sound). It doesn't take very much to alter our perceptions - a change in mood, exercise, tiredness, full or empty stomach, etc., may all affect how we perceive and how we rate our perception. The majority of these problems will not arise during comparisons over a short time.
Of course, a major difficulty with any comparison over a short timeframe is that, apart from not being able to assess our 'overall enjoyment' of the sound in a thoroughly relaxed state, we generally compare our current system (to which we are habituated) with a new one, so the problem of being habituated to the old system but not the new arises
I believe a technique such tune-dem gives the best chance of achieving useful short timeframe comparisons that are minimally affected by habituation effects, because it is directed less at the emotional response and more at how well we can make out the tune, instruments, etc. I agree that ultimately it is the emotional response that is the objective, but for the reasons above, I don't believe that we can reliably judge between two systems on this basis on either short or long timescales - however I do believe that the system that rates best for tune-dem is more likely to be the system that provides greatest enjoyment long term.
These are arguments based on little more than a smattering of theory, logic and vaguely related experiment (my degree thesis was on physiological response and habituation to environmental noise), and I'm making informed guesses as to the relevant timeframes over which habituation to the sound of a hifi system will occur, so it's all open to question, but it's the best I can do for now