Code factoring duplicates a lot of work

[colmsnowplow]

We can optionally configure behaviour for duplicates generation in the config.

The main function pattern matches whether or not this config is provided, and then calls a different function if the config is present vs if it isn't.

These functions do mostly the same things, but they're propagated through the entire codebase, everything is defined twice.

Is there a good reason not to just set the default values for the duplicates config all to 0, and use just one function?

If there is, we could still remove duplicate code by factoring it such that we create single events, then evaluate the duplicates behaviour and either a) produce the same event again, or b) re-generate data for the relevant fields in the event, and produce it again (depending on whether we want natural or synthetic dupe).