I'm not sure if I can explain this for you and make things better or just muddy the water more but I"ll give it a shot.
First your assesment of the control system is correct in a perfect world however we haven't taken into consideration "static droop". The basic principal says that the engine will do what ever it can to acheive an N2 datum vs actual N2 match. The datum is whatever value we decide N2 should be, in this case 107%. In real life however what the FCU is going to try to do is match centrifugal force against a spring pressure. The centrifugal force is coming from a flyweight that is driven by N2 and the spring pressure is set by the anticipator cable (connected to your collective lever). As you raise the collective you increase the spring tension and upset the balance which causes the FCU to increase fuel flow instead of waiting for the N2 to droop. If we didn't have this system in place you would have a "static droop" where the N2 rpm could be 107% on the ground but as you add load you would see a droop in N2. You would basically be behind the power curve all the time. The governor will increase fuel flow but not enough to maintain the RPM.
So in twin engine operation the system is setup to maintain a nominal N2 rpm (107%) by increasing fuel flow by a given amount to each engine as you raise the collective. Now when you stow one stove your only adding half of that additional fuel flow for the same collective position and hence you see a bit of a static droop effect.
If the system worked like the FADEC controlled aircraft you should not see this effect as the governor system simply adjusts fuel flow to maintain the N2 RPM datum. We still have the anticipator system in FADEC engines to reduce the transient droop. In the mechanical system the FCU doesn't care what speed the rotor or the N2 are at...only the centrifugal force that those speeds create and how it balances the anticipator spring tension. So at an OEI hover we have the same collective pitch postion adding an additional fuel flow to only one engine which will help carry the load but not quite enough to keep the same RPM as if it was operating along side it's partner engine.
This is much easier to explain if you have a diagram in front of you to see how the whole mess works. At any rate, if this explaination helped then good "I"ll be here till Tuesday!" and if it didn't then disregard all the above and listen to someone who can explain it better!!
Max