Aggregate design impacting performance?

Good day to everyone!
We are currently working with Axon Framework 4.9.* and need to migrate some data from a different database for one of our @Aggregates. Our current strategy is to reuse the same @Saga logic that is employed for the persistence of a new instance of the said aggregate via the user interface.
Using VisualVM to profile the performance of this migration process for around 200 aggregates of this type, I observed that this operation is using the most CPU:

Upon inspecting the code, I see that this operation is called when Domain Events are propagated from the @Aggregate root to its @AggregateMember(s). I must mention that there are indeed a lot of members in this aggregate, around ~15 HashMaps, and we do use ForwardMatchingInstances/@EntityId combination for event propagation.

What I would like to get more insight for is whether reducing the number of members in the aggregate will result in faster event processing types, event though most of these HashMaps are empty during migration.