If there is any sound difference due to processor use, would this not be shown up by encoding a track at different compression levels and comparing the result?
My recollection of Flac was that decode loading is virtually the same, it is the encode that loads the processor, but I may be mistaken Ali...