For those of you wondering about the effects of both better caps AND burn-in, I propose a simple and low cost test:
1) Buy an inexpensive but decent quality tube amp in its stock configuration. My example will be the Almarro A205A. Low power but very nice sound, even stock.
2) Listen to your reference recordings and get the amps strengths and weaknesses firmly in your auditory mind, as best you can.
2) Change the cheap output caps for something nicer. I put in Sonicap Platinums; lots of people use V-Caps, sometimes Mundorf, pick your flavor. Make no other modifications to the amp. Doing this yourself is usually pretty easy and keeps the cost down.
3) Put the amp back in your system and fire it up, no prep, no burn-in.
4) Put your reference recordings back on and listen to them again, once again noting the strengths and weaknesses.
5) Listen to the reference recordings again at 50, 100, 200 and 400 hours. Feel free to accelerate the burn-in by leaving the equipment on 24/7 until it reaches the hour checkpoints.
I predict that during 4) you'll hear immediate and noticeable improvement in many areas. This often the case even with far more expensive amps, because so many use parts that aren't very good. This has little to with good design or bad design; it's mostly about hitting a price point. If that isn't convincing that better caps are generally an improvement, I'm not sure what else will.
I further predict that at the checkpoints in 5) you'll hear noticeable differences, some that will be improvements, some that won't be, but that at all the checkpoints the sound will be converging on what you hear at 400 hours. The total number of hours to "settle in" will, of course, vary with the equipment and the caps, but you get the picture.
At 400 (or whatever your "settled in" point is), I predict you'll think the sound is improved, possibly greatly improved, from what you heard immediately after the upgrade. If that isn't convincing that burn-in is real and generally an improvement, I'm not sure what else will.
If you only bought the amp for the experiment and not to keep, chances are good you'll be able to re-sell it for $50-100 more than you paid for it, because lots of people believe doing such upgrades makes a difference and are will to pay for a piece where they're already done and burnt-in. At worst, you'll sell it for what you paid for it and the only cost of the experiment will be the caps and maybe some shipping.
I was skeptical about both better caps and burn-in effects until I did the above experiment. I am no longer a skeptic.
David