here are some "generic" guidelines about interconnects based on my experiences. Obviously, there are always exceptions to the "rules" : )
Smaller gauge wires typically present a leaner tonal balance.
Heavier gauge wires typically present a warmer tonal balance.
Solid wires typically tend to present a smoother, more cohesive presentation.
Stranded wires typically tend to present a brighter presentation.
Copper wires typically tend to present a fuller sound with greater output in the warmth region. Copper adds "body".
Silver wires typically tend to present a brighter sound with greater output in the upper mids / lower treble. Silver adds "detail".
Using the exact same materials in various designs with different geometries ( braiding, coaxial, twisted pairs or quads, etc.. ) can produce audibly different results.
High capacitance cables tend to roll-off the top end, reduce sibilance, add warmth, etc...
All cables sound better after being run on a Mobie or cable burning device of similar design for an extended period of time.
The key to making a "good" interconnect is to make one that is as transparent as possible that is resistive to RFI & EMI. To do this, the cable would not have any specific sonic traits and use geometries that are known to reduce inductance to reasonable levels.
To achieve these goals, one could take different types of wire combinations ( solid and stranded, copper and / or silver, thick and thin ) and go to town on various geometries. Since there wouldn't be any one dominating factor to either wire used in the circuit, all of the "resonances" would be spread out due to disimilar qualities. You would not hear any specific traits of either wire singled out.
Jon Risch's SSTP ( solid stranded twisted pair ) design using the cores from Belden 89259 & 89248 does just that. One wire is smaller gauge stranded copper, the other is larger gauge solid copper. Both cables make use of Teflon insulation for low dielectric absorption.
By playing with different combo's and geometries, you can literally "fine tune" various degrees of warmth, detail, air, liquidity, etc... once you get experienced. Obviously, the results might vary from system to system and component to component but you should have "somewhat predictable" results as you progress.
This can all be done for pennies on the dollar compared to commercial designs while resulting in a LOT of fun and education along the way. I highly encourage folks to try their hand at something like this. Sean