I'm not sure I agree that C isn't involved. And because the two leads carry signals that are 180 out of phase, the physically opposed coils would add, rather than subtract Z.
If you carefully examine the picture you see that the twisted cables are connected in parallel, both making the same link from the top of the ribbon to one of the transformer's input tabs.
Only for magnetically introduced hum they carry an out of phase signal (which cancels out at the transformer input), as the two coils formed between ribbon and either wire are running in opposite rotation.
This is the reason they did it that way.
Regarding stray capacitance: the ribbon is a signal source close to zero Ohms. Any C having an effect in the audio range would need to have a size in the uF range. This is by far not possible using some wires that short.
If you assume 0.1 Ohm for the ribbon you would need a 83 uF cap load to effect 20 kHz, equaling about 1000 km of twisted pairs.
The site you are citing is about transmission lines (e.g telefone lines of several km's lenght or Giga-Hz frequencies of network connections), and ONLY for those.
The loss effects described cannot be observed in normal audio range - studio distance situations.
Not to mention some cm's of wire inside a mic.
Looking at audio studio interconnects as transmission lines is a widely spread misconception.
Even then people usually don't follow the rules for transmission lines, e.g. proper termination.
They're just being sold over-expensive cables.