Nothing on there addresses the bottlenecks we saw during the last few stress tests. Why bump up block size to 128 MB (or at all) when it's clear we can't safely make 32 MB blocks yet?
Your argument dismantles itself, because then you could say we definitely don't need CTOR because it's just premature optimization not needed yet.
When did I make an argument? Or even mention CTOR? Im just asking a question, and pointing out the obvious.
