Physical explanation of amp's break in?

Sean...I doubt that the normal changes over time/use of resistor and capacitor values can explain what you claim to hear. These changes are trivial compared with the value tolerances of the components (2% is typical for resistors and 20% is common for capacitors). Circuits are designed with such tolerances in mind. Also, if values change, why should they preferentially change in such a direction as to improve audio performance?

Now, about that 6" wire that went to 50 ohms...you know that this wasn't normal It had broken strands or the like.
If it was solid wire, and it wasn't the terminations that were bad, I hope you sent the wire to some research agency for analysis. You discovered the "Sean effect"!

There are obvious reasons for speakers and phono pickups to change with age/use, so those items ought not to be a part of this great ongoing breakin debate.

Eldartford, your logic is flawed. Even though it is true that components are rated at some value plus/minus some percentage, it does not follow that a different value will sound the same even though it is within the rated specification. For example, if you built a crossover with all caps that are + 20% and another with all caps that are minus 20%, they would have decidely different cutoff frequencies and they would sound different.

As to the original weaker bass, stronger bass question, this could have something to do with the capacitors in the amp. The dielectric does change over time as voltage is applied to it. This affects the capacitance as well as the effective series resistance. I have no idea what the topology of this amp is or what type of caps are used, but I would not rule this out. Sean has given sound scientific explanations for his response, Bomarc, I would be curious to hear the reasoning behind your "nope."

Herman....Capacitors that are in frequency-sensitive circuits like a crossover, are selected to tighter tolerances, like 1%. In the usual power amplifier interstage coupling application, capacitors are specd to a much higher value than necessary so as to pass the lowest frequency of interest, so that even a 20% low item will be perfectly acceptable.

If you believe that the circuit is sensitive to exact values of components (that would not be a good design) the difference from unit to unit would be much greater than any variation of a particular unit over time and use.

And how about my question: How do the electronic components know which way to change so as to improve, rather than degrade, performance.

Herman, can initial change of new components be predictable? Meaning, for example, that for a specific type of capacitor, it's value goes down 0.2-0.4% on first X hours of use?

Eldartford, you are correct. Components are ignorant, but designers are not. They listen to their creations after everything has settled in, if it sounds worse than when they began they attempt to find out why. This is the same reason amps, etc. sound better when they are warmed up, that is how the designers listened to it when they finalized the design.

And yes, the circuits are sensitive to the value of componenets. Maybe not "exact" values, but there has to be some point at which the change is audible. Otherwise there would only be one value of resistor, capacitor, etc. Why does Levinson use laser trimmed resistors in some of their circuits? Because they believe that the exact value at that point in the circuit is critical.

Dmitrydr, I am not a capacitor engineer so I don't speak with great authority on how capacitors change over time, but I don't find it unreasonable to assume that if one was to apply a certain signal to a certain material that it's reaction would be predictable. So I would say yes to your question.