Edit

That is correct because it is already persisted to the DB. This is because the deletion of the misapplied name is a service call.

So this is an other issue for our save strategy ticket (please link). Because this is unclear for the user who expects that he/she can undo the delete. Also I wonder if this results in a flush which saves the accepted taxon (havn't checked this). If yes, the user needs to be warned. In long term we may want to do the delete in the same way as in the bulk editor now.