If you want to re-use similar elements across different content items you may provide your editors with a cloning functionality of nodes, paragraphs, media items, blocks etc. When you clone entities it's important that you always change their ID and UUID. Otherwise you may see updates of a cloned block appear on the original site after the content with the cloned entity has been pushed.
By default, Content Sync embeds blocks, paragraphs etc. into the content you are pushing. If you are re-using the same IDs or UUIDs for multiple content items then they are expected to have the same publishable content. Otherwise we can not reliably de-reference entities on the pulling site as we can't expect the same IDs and can only map by UUID. Drupal only provides UUIDs per entity and not per revision so that's a technical limitation we have to work with.
The Drupal revision system is not setup to track content changes in multiple dimensions (like git for example); it's meant to track linear changes of the same content over time. If you create a "basic page" you wouldn't expect two published pages of different revisions at different places for the same page because that's not what the revision system was designed for. If you want two completely independent entities through cloning then technically they must have different IDs (and more importantly for us different UUIDs). So cloning entities is perfectly fine but the cloned entity and the source must be independent and not share the same ID or UUID.