There's no "preferred" way, as such, but I guess if you're using the wrapper to work with the entity, it makes sense to stick with that convention to re-access it.
Don't forget you've just altered an in-memory object, so regardless of the persistence layer, that object will still have the changes in it.
Once you've done this:
Your changes will immediately be available by inspecting
EMWs use the core entity API under the hood, so persistence is actually done through the default entity controller. As such, everything that would happen in calling
node_save(), for example, happens when you update the wrapper object.
In practical terms, that means you can save an entity with EMW, and also access the changes on it immediately using an object returned from
entity_load(). It's the same underlying object in memory.
I think it's safe to say that context, and even surrounding code/conventions, will often dictate the particular method you use.