A heavy forwarder is presented in Splunk docs and other sources as being a type of forwarder that sends data to another Splunk Enterprise instance or to a third-party system. It is also suggested that a heavy forwarder can have a smaller footprint than a Splunk Enterprise indexer (by disabling some services such as Splunk Web), but retains most of the capabilities of an indexer. Finally, the docs will state that a heavy forwarder parses data before forwarding it and can route data based on criteria such as a source or type of event, and that it can index data locally while forwarding data to another indexer. However, you won't find a download file for a heavy forwarder on the Splunk site, nor will you find a specific document or direct references to setting up a heavy forwarder.
I'll offer another perspective on the subject of heavy forwarders...