Re: pex calibre view vs calibre view comparasion
Recommended to do a calibre NoRC vs. schematic comparison, with simulation. Strange things could come up there also and can help to find a reason. I don't know what were your settings, there are a lot of options when you generate extraction. The some bits difference can be a small difference in the log (date, name, target cell changed) maybe. These situations when people have to remember what could be else a month ago are such hopeless.
A possible solution maybe to create netlists with the 2 calibre views and compare the texts with a text comparison software. Unfortunately I can't say one, but I know it is possible and necessary sometimes.