I have seen at least one academic paper on this subject. My recollection of one of the papers is that it modeled the neck dynamics and envelope response of a rubber latex balloon for the same type of application as yours, namely a tightly coupled camera stabilizer system that used the latex balloon as the resisting body. Unfortunately I didn't keep a record of it.
I don't have the data to derive the change in angular momentum from an ascending latex balloon. But I can tell you that the overall system dynamics changes greatly based on the ascent velocity. The slower the ascent rate, the less drag created turbulence and thus the less motion present in the balloon system.