Software economics and code quality

So, I heard back from the NRC after I asked about software quality, start-up success and CMMI. They got back to me very quickly with a good nugget of information;

I just have one comment… using CMM-based assessment as a proxy for quality is not a good idea. CMMI is a process model, and as such does not really address product quality (it just assumes that heavy process = quality, and advocates measurement).

So I continue my search for the truth about start-up success and software quality. I just finished re-reading a great paper named “Big Ball of Mud“. The paper is all about softwares defacto implementation architecture. What i have gleaned from this paper is “use an XP style methodology and refactor, refactor, refactor.”