The original note, as written at the time
The player doesn’t have more than 50 percent control of all things happening in the game. He can direct the game towards a particular theme of outcome but the details are not for him to decide. At each key milestone the player must take on a detail role. Say he launches a major spy operation — during the next 30 or so turns he actually participates as the spy. The gamer is like a spirit to the characters in the game.
Ongoing development
A game in which the player never has more than half the control. They can steer the outcome towards a theme, but the details are not theirs. At each key milestone they must take on a role — launch a spy operation and for the next thirty turns you are the spy.
The design is unusual because it treats the player as a spirit rather than a hand. In most strategy games you are an omnipotent cursor; here you are an influence with a limited budget of direct intervention, and the rest of the world moves on its own.
The role-taking is the good part. It forces the player to live inside the consequences of the decision they made from above. You ordered the operation; now you are the person in the corridor with a radio, and the corridor is nothing like the map.
That switch is what most strategy games are missing. A player who only ever commands never learns what a command costs, because the units are counters. Being the counter for thirty turns teaches it in a way no statistic can.
The risk is friction. A player who wanted a strategy game may resent being dropped into a stealth section, and a player who wanted the stealth section will resent the map. So the transitions have to be short, legible, and skippable at a price.
The deeper idea is worth keeping even in a smaller game: the commander is not the only point of view, and a game that lets you see from both is a game about responsibility rather than optimisation.
Give the player power, then make them live in it.