My approach is to identify and model exactly what each coil is doing and optimize toward the ideal behavior. That is different than choosing which sound I might prefer. Coils are quite technical in how their electromagnetic fields and therefore transient behavior develops. A primary source of distortion comes from wire squirm which smears the signal. Thiel used a dipped and baked coil round wire. Foil further reduces mechanical squirm as well as develops tighter field effects. They are better inductors with fewer side-effects. They also cost more. They also remove an (artificial) roundness and excitement caused by the euphonic effects noted above.
Due to practical conciderations, I am investigating replacing series feed inductors at 18 gauge and larger with foil. Smaller gauge foil equivalents are not readily available and Jim used 22 gauge coils in shunts specifically for the fine-tunable inductive resistance they provided. (Bigger is not always better.) Those resonance circuits do not carry current and are explicitly tuned to driver and overall circuit parameters. Besides, they are world-class high purity copper in custom dimensions and oven-baked. Hard to improve.
Due to practical conciderations, I am investigating replacing series feed inductors at 18 gauge and larger with foil. Smaller gauge foil equivalents are not readily available and Jim used 22 gauge coils in shunts specifically for the fine-tunable inductive resistance they provided. (Bigger is not always better.) Those resonance circuits do not carry current and are explicitly tuned to driver and overall circuit parameters. Besides, they are world-class high purity copper in custom dimensions and oven-baked. Hard to improve.