There is also the possibility that rotation inertia plays a roll; and there is the hypothesis of John Middleditch that virtually all supernovae a binary events that are differentiated primarily by viewing angle: New paper:
http://xxx.lanl.gov/PS_cache/arxiv/p...708.2263v1.pdf
The mainstream position remains that carefully characterized type Ia supernova are a special class of events in which the absolute magnitude is a function of strick mass and compositional boundaries. The observations noted in this thread do not support this simplistic approach in using supernovae to measure cosmic distance. Multiple parameterics come into play.