>why is one decomposition favored

In analog electronics, sine waves are everywhere due to the way
current and voltage behave in analog circuits. Building a filter in
this environment using a trig decomposition is easier than switching
to another decomposition.

In math, sine waves form a nice orthogonal basis, which makes the
decomposition easier to calculate.

In the ear, natural vibrations are probably best modeled with sines
and cosines. In other words, fewer basis functions are required to
describe the vibration when the trigonometric basis is used.

It seems that we want to believe that the mechanism of hearing
involves a decomposition into a trigonometric basis followed by a
cognitive synthesis. I'm not sure why this would be, and what this
model accomplishes, other than satisfying our urge for reductive
analysis armed with the most convenient mathematical language known to
us.

Doug