I'm not sure there is a direct question in there for me to respond to, so all I'll say is that it is a minimal up-front effort strategy that puts the hooks in place that would allow some real measurements to be taken.
And even that minimal effort needn't be discarded if the LLVM experiments fail, because it leaves an intact perl distribution that works just as efficiently and effectively as it does now, with the only difference being that the compiled perl executable comes in 3 parts rather than the current two. (In windows terms: perl.exe dynalinked to perl5.x.part1&2.dll & perl5.x.part3.dll)
And that split might even benefit the normal distribution in some way.
Once the hooks were in there, the ability to pass the PCG to some other dll, module or process, might allow other investigations to move forward.
With the rise and rise of 'Social' network sites: 'Computers are making people easier to use everyday'
Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
"Science is about questioning the status quo. Questioning authority".
In the absence of evidence, opinion is indistinguishable from prejudice.