I don't recall the exact discussion from P2D2, but AFAIK the reasoning was more along the lines "There are other bottlenecks that we need to address first." Tha

I don't recall the exact discussion from P2D2, but AFAIK the reasoning was more along the lines "There are other bottlenecks that we need to address first." That is, issues that would either limit the JIT gains or issues with better cost/benefit ratio (measured in developer time).