I agree that skin factor may be an important effect to be considered in choosing correct conductors for a specific application. However I don't see how this has anything to do with the question of whether a stranded or unstranded conductor of equal DC current carrying capacity is or isn't better for cases where skin effect IS significant.

Skin effect is a function of the signal properties, not the conductor properties.

What am I missing?

 

True... but when selecting a conductor for a given application it's the load's properties that should guide the decision. For a headlight, the only reason to use stranded wire is to take advantage of its greater reliability under mechanical stress. For a more complex load that has high frequency components, we use stranded wire to provide a greater surface area for any given cross-sectional area, thus reducing transmission line impedance at those higher frequencies. Due to skin effect.

Just the high frequency components of the current drawn by certain loads, such as inverters which tend to gulp current in (relatively) square pulses. The basic definition of a square wave is that it's a signal consisting of an infinite number of odd harmonics of the fundamental frequency, so there's just no escaping the high frequency components. So when we minimize the source and transmission line impedance at higher frequencies, we get sharper current waveform rise times which leads to cooler semiconductors (by getting those switching transistors out of their linear regions faster) and improved system efficiency due to reduced transmission line loss. A good inverter will also show faster transient response, but most cheap consumer-grade crap has the transient response time of gelatin anyway.

 

I'm right with the concept that: as frequency of a signal rises it's current density moves from the center of a conductor outward along the radius, and insufficient conductor radius for a given frequency will result in a higher impedance. Also right with you when you say the transients of a switched power supply, such as a modern DC -> AC converter, are typically far from being merely DC spectrally.

I still don't see where surface area comes in. While the wiki link is describing cylindrical (presumably solid core) conductor skin effect behavior, it seemingly completely neglects to address the stranded case. Are there other formulas not found at the wiki link you are using that show this relationship?