@jtimon It’s not so much about ‘review costs’, because review will almost certainly not find all bugs/inconsistencies. It is the costs of some unexpected change causing a fork. Years after Satoshi left people are still finding small unexpected cases in the consensus code.
Just that no one can point them out from glancing at the code doesn’t mean that no problems exist. Making the code (maybe) a bit more readable and consistent is just not worth that. I hope you understand. If the potential benefits don’t outweigh the potential trouble no one wants to take the risk to merge it, and a NACK is in order. That’s better than keeping it in limbo forever, IMO.
@leofidus Yes if this could be done only outside the consensus code, it would be acceptable. That’d require an interface layer between the consensus code and the rest, which would be a good thing anyway. It is aligned with the goal of making a consensus library.