Just an opinion. I increasingly favor the double degeneracy model for SN Ia. The occurence frequency looks about right, and we have hardly any confirmed companion stars for most SN Ia. While most SN Ia are too distant to have much chance of detecting a companion star, it is also true of [the admittedly few] nearby SN Ia over the last millenium. One problem I see with double degeneracy is energy released appears less constant than in the accretion model. On the other hand, we have plenty of irregular SN Ia, which suggests multiple modes. That seems reasonable. Curious what others think.