It's not that simple. Like I said in another comment, organising the data is only half the battle. This optimisation also depends on your code accessing the dat

It's not that simple. Like I said in another comment, organising the data is only half the battle. This optimisation also depends on your code accessing the data in an optimal way. If the code itself is not organised for DOD, then the "optimal" organisation is what we currently have. A per-entity Update method that's accessing different "kinds of data" will perform worse with DOD-organised data. This is why we have an architectural pattern that automates all that, called ECS.