I ran a very simple simulation using a single voltage source with ideal 0.1-ohm output resistance, an ideal 8-ohm load, and a series resistance of another 0.1 ohms using a SPICE H-source and series 0-V voltage source to emulate a resistor whose value changes (to avoid "is modulated by") the current through it. This is a common way to model an ideal
noiseless resistor in simulation so is something I've done "forever" when I need such a thing (for instance to get circuit noise without source/load resistors).
Schematic:
View attachment 62182
Result with H1 (ideal resistor) set to fixed 0.1 ohms (spurs are high because I did not bother trying to tighten tolerances and such):
View attachment 62183
Repeated with H1 changing with current to model a simple fuse'ish, no time constant/cap, just to show what can happen -- notice how the second harmonic has popped up:
View attachment 62187
This is not a real-world situation but is just to show that a series resistance changing (varying, "modulating") with signal current can add distortion.
HTH - Don