I am an audiologist who programs for a living (audio-stuff, obviously). I guess that must make me a really relaxed personality. Only I don't live in the US. Oh well.
To a certain extend, yes. But then again, it is not so much about mathematic accuracy as it is about 'sounding good'. Many of the most successful audio effect devices have really meager processors. Careful tuning of simple algorithms is often more productive than crazy mathematical forays.
(This is kind of the same thing that applies to Openoffice and MS Office: OOo is not really technically any worse than MSO, but MSO just has the nicer presets and that makes all the difference).
That said, you are often limited by available processing power and micro-optimizations can be important. Calculating dozens of filters for dozens of channels on one DSP can be quite a challenge and getting that right algorithmically makes a big difference.
