Why do you think Bi-Wiring improves the sound ?

In my experience, my Silverline Sonata-III speakers with biwired cable sound more analytical, even if the cable is warm (like the Silverline's own biwire cable).  

I own bi-wireable speakers. I also had a hard time accepting that
splitting one cable would possibly make a difference. At that time I was
using some heavy gauge, copper Monster wire and planned an upgrade of some kind.
I was in the Kimber Room at a show in 2019 and asked "Who here might best counsel be on which Kimber cable to buy"? The gentleman I asked said "I think I can do that".
After asking about system and my listening style he suggested
an 8 biwire product. A 3 meter run was about $280.00 and could be made and delivered in 3-4 days.  
The 8 wire strand would be split as follows: 3 to treble, 5 to bass.

This was Mr Ray Kimber himself assisting me.

Interesting side note: My counselor/mentor/friend who has the same speakers as I do was recently listening at my home. He was trying to figure out why my system produced better sound than his. He does not bi-wire.
So I feel I have an improvement thanks to the bi-wire.
Did I do a blind test? No. 

Has anyone got some scientific blind test results??

Oddly my girlfriend and I just spoke about this last week while she was painting my fingernails.  Coronavirus social distancing is getting to her.
Digressing.  We have this Plate Lunch equivalent purchase values.  She freaked when I bought MIT biwire cables.  We could have eaten takeout twice a week for a year on what I spent on them.  However, when I installed them, she immediately heard the difference (good) when she walked into the room.  Not even in the sweet spot.  She does have a really good ear.  So we spoke about it. Why?  I gave her my 40 years in electronics reasons why it should  not make a difference.  But it does.  She came up with this ordering take out analogy.  I order shrimp scampi and she orders a caesar salad.  The restaurant puts them both into ONE container (one cable).   DoorDash delivers, she will have scampi on her salad and I'm going to have caesar dressing on my scampi.  That would not happen if they are placed into two different containers.  And the longer the trip, the greater the contamination.
That's a blonde non audiophile thought.  And I now have red fingernails.

I always liked this article, for what its worth.....

https://www.shop.us.kef.com/blog/a-quick-guide-to-bi-wiring

Alan Shaw, the designer of Harbeth speakers, has typically strong views on the topic. These three quotes were taken from three separate posts on the Harbeth forum.

"I really wish this subject of biwiring would just disappear up its own terminals. I don't have many ambitions in life but killing this discussion by deleting the biwire terminals and reverting to just one input pair is going to be at the top of my 2011 New Year's Resolution list for the remaining models that still feature biwire legacy terminals! You've given me a real motivational boost!

"The terminals were fitted for one reason and one reason only: to give the user choice. Have I ever used them at exhibitions? No. Have I ever used them for critical listening? No. Have I ever used them during the design of the speaker? No. When we were offering the biwire terminals, right at the end of the design process (which has all been with single wire) I took a saw to the prototype PCB, cut in in half to isolate the bass and tweeter sections and then made a pretty PCB layout based on that. Did I listen to the biwired crossover before authorising production? No. Do I believe that even 0.00000001% of enhanced performance can be gained? No. 

"Of all the subjects ranged over in the speaker arena, this one is a complete and utter waste of time - in my opinion. But what do I know about it? I only design the speakers ....... !

"Biwiring does do one thing very well though: it introduces the one and only, much appreciated 'fiddle factor' to allow individuals a physical and psychological interaction with their speakers. What else can you do to them other than dust and polish them?"

***

"A biwire link is gold plated brass about, say, 30mm long. The claim is that this particular 30mm long piece of highly conductive metal is somehow, magically, more important than any other 30mm piece of perhaps less highly conductive metal anywhere else in the chain between the loudspeaker drive units and the power station a hundred miles away which is supplying the current that causes the cone to move and a sound to be generated. Does that sound logical? Does that sound an intellectual argument that a professor of engineering at a university could or should set his students studying? Of course not. It's a daft fixation on what is, from a point of electrical conduction, probably the best "link" in the chain from the point that the mains supply enters the house.

"The biwire link has this fascination for one reason and one reason alone - it's accessible by the user. So it lends itself to being fiddled with and to all the associated gratification of adjusting ones hifi.

"This is a non-issue. Pick a genuinely 'weak' part of the signal chain and experiment, but this big, fat brass part with countless billions of surplus electrons isn't the hold grail. Of that I am totally and absolutely certain as I've stated. You'd be better off paying attention to, let's say, 30mm of copper track on the printed circuit board that the binding posts are pressing onto which is vastly less conductive because it is thousands or millions of times thinner than the biwire link. But of course, that would involve opening the speaker and voiding the warranty."

***

"Do the exponents of this biwire mania have any concept at all that a current is a circulating concept? Circulating from the power station, through your amp, cables, crossover, voice coil and back again to the power station? Anyone into biwire connectors grasping that concept please? That concept of how electricity actually works is why there is a live and neutral pin on your wall socket. There has to be a flow. And what impedes the flow is resistance. And resistance is associated with thin parts, like the voice coil (about 6 ohms). So the fact that the biwire link has a resistance of perhaps 0.000001 ohm compared to the voice coil's 6 ohms means that as a component in the circulating loop, what dominates the resistance by a huge factor is the voice coil.

"If the concept of a circulating current is unclear or distrusted then the whole scientific world we live in collapses."