I don't have scientific proof specifically, just a lot of listening and testing experience. I would NOT splice particularly using adapters to put two shorter pieces together. I don't think people appreciate the importance of speaker cable in case of low impedance speakers with extreme impedance swing MATCHING with power amps with high damping factor( low output impedance). I just built the low inductance cable like you linked http://diyaudioprojects.com/Power/Lo...peaker-Cables/
But using 6 pairs of 16 gauge. I actually disassembled one of my cable on ONE SIDE and just redo the twisting without changing the cable. It sounds quite different. Same length of cable, same connector, just different twisting. It is that important. I just did in on Friday. I did a DBT and people notice the difference on Saturday. It's very obvious.
I had built speaker/amp switching fixture that I can switch different amps to drive the same pair of speakers, driving by the same music source. I took precaution to use very heavy duty relays, very short and thick wires in the switching fixture. I was hoping that it will not degrade the sound. I made short low inductance speaker cables to connect the switching fixture. It FAILED
, it degraded the sound. That was a total waste of time and money. I still have to hand switch speaker cables to switch amps. I so wish the switching fixture would work. It would be so easy as I can just sit in the chain at optimal position, flip the switch and compare two different amps. That I can have people listen without knowing which amp is using. BUT IT FAILED as it degrade the sound.
I am NOT an advocate in spending $$$ on cables, I build my own, it's $30 a pair. But my experience is it's very critical to make the cable as one complete cable with minimal connectors in between. Using just enough length for the setup.
I want to specify that what I said applies ONLY to speakers that are low impedance with extreme impedance swing MATCHING
with power amp with very high damping factor. For tube amps that has damping factor like 1 to 5, I am sure you can use any cable and have any splicing in between will not matter.
I explained a few times, it's NOT the loss due to the reactance of the cable inductance as we agree that the reactance is in the range of 0.3ohm at 20KHz. IT's the THD increase due to the reactance of the cable inductance. I actually have a thread here on banana plug vs spade connector.