Bass leaves after amp warms up?

"But why didn't it end a long time ago? "

I could be wrong but I do not recall the ill advised attempt to bi-wire the speakers from a single source amplifier ever coming up until late in the game, after the issue had apparently been addressed already. That would have been a good red flag early on. Not sure how/why that did not come up earlier? In lieu of that, I think we were all grasping at straws in lieu of that key fact.

Any lessons learned? What might have brought this to a happy conclusion quicker?

Even now, I am not quite sure the bi-wiring was done in a harmful way, but it was clearly ill advised and increased the chances of doing it wrong and experiencing problems as a result.

There was so much going on at the OPs end that it was very hard to keep track of what the current state of his system at any particular time was. Not a good case study for how to efficiently and systematically resolve a problem with ones stereo, though still an interesting case study to learn from nonetheless.

Its always hard to diagnose a problem that one (other than OP and his crew) cannot actually hear. Internet threads are always limited that way, not the most effective tool alone for sure. Its a game one plays by choice realizing the limitations I suppose but with the hope that things will work out still in the end faster than they might otherwise.

So, what did you learn?

For me it reminded me of the the need to make sure fundamentals (like proper wiring) are correct and not assume that just because someone does not like what they hear, that it is necessarily the equipment's fault. I knew all of that already I suppose but this was a good reminder to sometimes expect the unexpected.

Another lesson learned for me was when something is not going right, keep things as simple as possible. For example, don't attempt to bi-wire right out of the gate. Audiophiles (including myself) tend to make things more complicated over time.

WHen something goes wrong, you have to isolate the problem. That's harder to do as the system configuration becomes more complex. The best approach may be to stip things back down to a minimum, and regression test from there. That may still often require having spare pieces around to swap in and out for comparison when needed.