Personally I've experienced a bell-curve of expectations from conflict systems, from simple to complex to simple again. It's been an evolutionary path that mirrors the kind of stories I've liked and like to create: starting with simplistic goals ("wack the bad guy, loot the treasure chest.") have changed to complex explorations on _how_ goals are met, and finally an epiphany that _how_ isn't as important and _why_ and _whats the outcome_ of perusing goals and change.
In my youth, I sought out detailed, complicated and 'crunchy' game systems because I sought understanding how the things worked, even in make believe worlds. The kind of games where a six-second combat round could take half an hour of real time doing math and looking up rules and charts. But in my experience it's only fun when it's warranted. As a GM it's a nightmare to track tons of details and characters and plot.
In other words, I'm becoming and old phart and don't want to consult a ton of charts and use my slide-rule in pursuit of a good story anymore.
Personally, I'd rather focus on the tactics and strategy and not get overwhelmed by the math. This was no more apparent to me when I heard that AV:T had been licensed as the board game engine for an Honor Harrington game of starship combat. The novels focus on the leadership and tactics of the ship captains, not so much the navigators' working sliderules and working graphing calculators.
Then there's the other end of the spectrum; games like Full Thrust and Battlefleet Gothic, which are almost too simple in their rulesets, but let players concentrate more on tactics and fleet composition than number crunching.
To delve deeper into the separation of player-vs.-character abilities: If my character is a brilliant engineer/programmer/scientist, then I the player shouldn't be limited by my relative lack of ability. The player states their intention, rolls the dice, and the GM and player cooperate to narrate the outcome. Mechanics should determine degree of success or failure, but leave the rest to narration.
Players should focus on choices. Mechanics should support choice.