But wireless audio systems almost always use companding to improve S/N ratio. Why bother unless maybe FM isn't quite as clean as we'd like it to be?
It can be as clean as we'd like, but it's not, because the FCC and most authorities, worldwide, limit the signals that we use to a maximum of 200KHz RF bandwidth. Smaller RF bandwidth means smaller maximum FM deviation, which means lower audio dynamic range and higher audio noise floor. Companding is the simplest means of mitigating that.
In RF, overall performance is determined by a balance of all variables. With unlimited spectrum and no regulation, current analog technology could easily produce nearly flawless audio and extremely long range, but this can sadly never be so because RF spectrum is a limited natural resource that many interests rely on and compete for. The development of digital RF technology is fueled by the same motivation that drives the development of wired communications; to deliver more (or better quality) content while using less bandwidth.
