We have some cases where we get data from multiple input topics for the same set of transformations. Given the fact that Candyfloss only allows a single input topic, this means we have to create multiple deployments which means we have the added requirements of additional resources for each instance. Having the ability to pull data from multiple topics would reduce the overhead and allow for more efficient deployments.