Read a couple, and have a couple ideas as to problems, such as feature creep (which I know by the name requirements volatility in terms of quantifiable attributes), poor integration of multiple asset pipelines/processes, lack of user communication or overabundance (lack of causes problem misinterpretation, current overabundance is usually due to prototyping, which leads to prototype as product), distributed work problems (time zones) and poor requirements. I do not know how Quality Assurance and Performance Engineering are tackled in game projects either, but I am guessing that for most (small) teams performance engineering is non-existent. Even with all this, my main problem is still the fact I can't prove these are actually the problem, since I have no tools with which to take measurements of these attributes.
Also, despite all this, are there any particulars you might recommend. I've read the GDD article on GUM, as well as Evolutionary Design. I am not counting forum posts since those were a mixed bag.
P.S.: I could also use any writings that deal with art or music as a stand-alone process to determine the differences from the standard software models (which tend to be code first). Also, I'll take a look at GDMag. Thanks.