There's two things we could try to optimize the PCR reaction that we haven't tried.
Add DMSO. In molbio folklore, often this is a "magic" reagent that helps straighten out DNA. Mechanistically not sure how it works, but I know empirically that it can help.
Try 40 cycles of PCR. That would equal to 210-fold amplification over 30 cycles.
@Starnite - just thought of something.
Checked page 10 of the Agilent enzyme manual. http://www.agilent.com/cs/library/usermanuals/public/200550.pdf
There's two things we could try to optimize the PCR reaction that we haven't tried.
What do you think?