This is not a bug and is more of an issue in design. See:
https://github.com/Thalassicus/cep-bnw/issues/224 and
http://forums.civfanatics.com/showthread.php?t=503813
I can adjust the amount of population lost and the resistance period for the Mercy and Force options so if you guys come up with a good alternative I can make the adjustments.
I don't see anything that in this discussion that ever proposed removing the city resistance mechanic. In fact, the design proposal specifically says that the city capture options are to have "normal gameplay effects".
http://forums.civfanatics.com/showpost.php?p=12685513&postcount=52
What I understood the intended design of the three features was:
a) Mercy = standard GEM style city capture with minimal population loss and not much building destruction.
You pay a high upfront price in terms of unhappiness, and have a long period of resistance.
But after that, you get a very profitable city, which still has a large population and lots of buildings.
b) Force = standard vanilla city capture, with major population loss (half) and more building destruction.
The upfront cost is smaller: you kill half the population and so have less unhappiness to worry about, and the resistance period is shorter (because resistance is a function of city population). And you get a larger gold income from capture (=the sale value of the destroyed buildings?).
But the long term benefit is smaller: the city has a smaller population and fewer buildings.
c) Wrath = standard raze
High upfront unhappiness, but zero long-term unhappiness or social policy/science penalties (because the city will no longer exist), and effective denial to your enemies if you don't think you can hold the city long term.
Mercy and Wrath would both have standard resistance, which IIRC is the size of the population after capture, though potentially we could have something with slightly less than this.
[eg Resistance goes up by 1 per pop up size 10, then 2 per 3 pop up to size 20, then 1 per 2 pop above that. So a size 16 city would have 14 turns of resistance. And then there could be social policy effects - especially in Autocracy - that reduce that. It could be less than this too, these values might be too high. 1 turn per pop up to size 5 and 1 per 2 pop after that would probably be plenty.]
