Good summary. We do know that KBLB is now using CRISPR and not the less consistent earlier insertion technologies, so most of the issues related DIRECTLY to the insertion should be considered as rare as mutations.
Too variable of a starting platform worm and/or inconsistent population monitoring make the most sense to me, especially if they were trying to cross breed their insertions into lines with new genetic backgrounds (VN worms, etc.)
That platform worm strain should have been much more carefully culled for outside variation, and as much of pain as it may be, if you want a VN-platform worm, you need to start from scratch and make a new one through controlled breeding.
My top guess is still that corners were cut with respect to maintenance of clean lines, early on in the ramp up or during the platform worm generation. The new testing setup will be helpful in choosing new starter colonies. I would absolutely couple that with a few rounds of genome sequencing. That's what a pro would do.
EDIT:
Keep in mind, the variation in silk quality must have been substantial. I highly doubt that Vietnamese silkworm farms have hermetically separated breeding setups, yet they crank out silk at production quality. The variation found in our silk is probably not subtle epistatic variation (IMHO). Some gene/allele that significantly affected silk production infiltrated the population. Where's the most likely source? Vague epistasis, or competing silk-production alleles. I vote the latter, but admit the former is possible.
Market Data 
Markets 
AI
