Calling setstakingreward with the same inputs will continue to append to the list,
which is unexpected.
The impact of fixing this alone is relatively low but sets the stage for refactoring and using
this code in other places, such as pre-consensus on staking rewards.