This is what I am working on right now, digging through the assembly of the 2% version to see what is wrong with this version. But honestly, whatever compiler he used has generated the instructions in a really weird fashion that is slowing me down.
Cool, you guys are way smarter than I am when it comes to this stuff. All I'm capable of doing is setting up a development environment in Ubuntu to compile from source lol. The only help I can offer is testing :-) Take your time. I'm sure you'll figure it out!