Eliminating spade connectors, upgrading bits, soldering all of it in

At issue with  connectors is connector quality and metallurgy. OEM often choose cost over quality, and then we have all these interactions with various metals coming into contact with one another. All my line, phono preamps and SET and push pull tube amps point to point wired, no more circuit boards in these components for me. I also don't solder every connection, don't like crimps but have some clamping connectors. Paying great attention to wire quality, connections, especially in signal path is as important as the caps, resistors, etc to me. The boutique parts manufacturers also find metallurgy to be critical in voicing of their products.

@decooney, you're a brave man posting such an incitive outrageous idea. I commend you for it.  For mass produced speakers on an assembly line a preassembled cable harness with press-on connectors speeds up the whole process, saving time and skilled labour and time as we all know is money. It most certainly is not for superior sound. They mostly consist of Brass and Tin or Nickel plating and all the metals are low down on the IACS chart.

There is also much to be gained by improving the XO in terms of orienting the inductors correctly (costs the modifier nothing except a little time) and replacing the ubiquitous white sand cast resistors most of which have magnetic ends.

Now before the swarm of deniers pile on the above statement I challenge you to a simple experiment: To those who are skeptic and have access to the XO inside their speakers but taking care not to void their warranty, try this:  In almost every case there will be at least one and usually two resistors forming an L-Pad to attenuate the signal to the tweeter, yes? Yes. Find the series resistor, look for that noisy white thing, often with a value between 2 and 6.8 ohms. Lets go with 6R8. Unsolder it and replace it with a string of 10 x 0.68ohms all soldered in series = 6R8 and listen to the difference and you will hear the damage components make.

Expectation bias be damned!  Consider too the big bold speaker binding posts even the $$$ speaker brands use.  Brass is used because Cu. does not machine as easily and strips under torque and then plated with Nickel because it polishes beautifully and then finally Gold plated resulting in a georgeous piece of audiophile porn. You now have a chunk of 3 dissimilr metals in your precious signal path.

Good speaker terminals are scarce but available. 

As @waytoomuchstuff ​​​​and @gdaddy1 in essence stated: It all adds up.

Shocking!  No one with really good equipment would do this.  Value just went to ZERO!!

@gournard In almost every case there will be at least one and usually two resistors forming an L-Pad to attenuate the signal to the tweeter, yes? Yes. Find the series resistor, look for that noisy white thing, often with a value between 2 and 6.8 ohms. Lets go with 6R8. Unsolder it and replace it with a string of 10 x 0.68ohms all soldered in series = 6R8 and listen to the difference and you will hear the damage components make.

Yep, 100%, exactly the point.  Over four decades, on the 43rd and 44th set of speakers I built, with some help from a few friends with access to an isobaric chamber, who helped me on the design of more capable crossovers - -  Once the best L-pad settings were determined, the prior L-pads in the circuit were completely removed and replaced with the right resistors and filters in the next pair of speakers.  Never looked back since. 

The next step was to compare an original set A with L-pads and the updated set B without L-pads.  The difference was quite notable. More open, detail, and more realism to instrumentation.  I often times use piano keys to judge difference in A/B test cases.   I A/B’d these two sets back and forth over two weeks, and there was no reason to continue with any L-pads in the speakers at all. Get ’em outta there. 

"Damage removed", out of the signal path, making a rewarding difference here too. One of the other benefits of building your own, DIY, revise the design how you see fit to your taste.  

Other-Beta Test:

During this same time I completed an evaluation for a known manufacturer, with  a 3rd set, completely stock with zero mods, new from the factory. These were intermixed with the two other pairs with minor to major mods, upgrades.  It did not take long for the 3rd set to get boxed up and sent back after their normal break-in time had passed.  No reason to keep them around, they placed #3 out of 3 pairs.   

When I changed my speaker wires from relatively inexpensive Synergistic Research wires to Silversmith Fidelium cables, the change was pretty dramatic. I talked to Jeffrey Smith about it and he suggested trying some heavy banana to spade adapters.  He explained what this did.  It made the Fidelium sound more like my old wires, but still better.  I ultimately got rid of the adapters, but I found the whole concept quite interesting.