Some thoughts here. Strategy games are interesting beasts… on the one hand, you have the actual strategy mechanics, and the strategy mechanics should provide for interesting decisions, which should have varying and non-obvious results in terms of your success in the game. If an optimal path is obvious through a strategy game, that’s a problem. Additionally, strategy mechanics imply variety – if the same choices are given to you every time, you’ll quickly learn which ones are optimal.
On the other hand, you have the immersion of the game subject. Board games generally provide minimal immersion – they’re mostly about the mechanics and about interacting with other people. The mechanics tend to be very abstractly related to the subject matter since you have a very low complexity budget (mortals are expected to run the game, after all). In order to be able to run the game within the limited complexity budget and still provide non-obvious choices, you have to have heavy abstraction, with the subject matter often serving only as inspiration.
Computer games can go much heavier on the complexity budget, since now the game is managed by a machine. This, together with improved presentation, can give a much better feeling of immersion than anything supplied by a board game. However, this leads to a different problem: because you can have so many mechanics interacting, it’s much harder to make sure that the experience generated for the player actually has meaningful choices (i.e. non-obviously optimal ones) to make, and that the experience is sufficiently varied. Additionally, the computerized game experience may include entire artificial enemies (AI), which can lead to serious problems, as a malfunctioning AI hurts both immersion and strategy, while a non-malfunctioning AI is extremely difficult to program.
Paradox games sit very heavily on the side of immersion. They are the ultimate computerized strategy games, running as they do in pausable realtime. They feel like simulations of entire virtual worlds, with their hundreds of AIs, and this is one of the reasons people love them so much. But they suffer from being so far down on the immersion scale: it’s very hard to make sure that real strategic decisions are provided to the player, rather than ‘busywork’ (obviously optimal actions) to make the game hum along. There are so many mechanics, in fact, and so many custom ones per race/nation, that they provide multiple unique immersive experiences, depending on which part of the gameplay you wish to sample. But those experiences are highly unlikely to be strategically challenging. In this way, they really do cross over to story generators - procedural RPGs where you’re just trying to play out a story - rather than fully functioning strategy games. All computerized strategy games possess this element of story generation, but the best ones have it balanced out by the strategy elements.