We skip fetching IDs that were just fetched a moment ago, so the array size will be reduced, but the reducer was still expecting all items to be fetched, hence the array size mismatch
1. The API supports batching -- updated the code to use that. Retained string as the parameter (instead of changing it to array) so that existing clients won't be affected.
2. Make `doFetchSubCount` a batched command by default through an idle timer. This way, none of the clients need to collect IDs -- it's all done behind the scenes.
3. Added minimum of 5 minutes between each sub-count fetch for a claim ID.