I think so too. You could also take it one step lighter and just store the aggregate itself temporally, as SQL Server will keep track of any mutations on the aggregate (those are your events!) and you could mold those history mutations into the respective events. It's not strict event sourcing, but may be an interesting road to go down.