I would probably enjoy that too. But note that the commander/player in such a game never actually sees the enemy; all they see is their own units taking off and landing, until the enemy attack planes arrive. There was no ship-on-ship combat at Midway, right?

The outcome of Midway was determined by an intelligence trick: a planted plaintext message about the condition of the desalination plant at Midway island. I wonder how you'd model that sort of thing. Also, there was an incident where an outbound US attack squadron eyeballed an incoming Japanese attack force. What are the chances of that, and how do you model it?

In general, modelling intelligence seems to be hard; I've neveer seen it tried, except in a very abstract way.