So your simplified system might cater to group one (which may or may not be the majority of players anyway), but will certainly drive away group 2. To them, the metagame and number crunching IS the game, playing the game and winning against less informed players is just collecting the price for their hard work.
I don't have numbers, but I think its probably safe to assume that if your game leans towards simulation then you want more, (and more detailed) choices because the crunchers will be a significant portion of the player base (I would contend though that crunchers are not, in fact, the majority even in any but the most hardcore/niche simulations). Looking at racing games, you have the likes of Forza or Gran Turismo or F1 that are deep simulators and the details are necessary -- still, Forza and Gran Turismo, and to a lesser degree F1, still have a majority audience who are not crunchers. Even if those non-crunchers make choices because they are forced, or because they think they should, they don't understand the choices they're making and will make choices that don't benefit them.
Arcade-style racing games go the other direction, with very few choices and configuration details. Things are not so granular as to be overwhelming, and I relate to each of the limited details in ways that mean something to me. I know, for instance, that if I'm not a great player of racing games I should look for a car with better turning radius and acceleration, even if I give up top-end-speed and breaking, because that will better suit my style of play or skill level -- I know that top-end speed won't be of use to me unless I'm a deft driver.
If you haven't looked at that link, one of the examples she actually used was in studying how customers engage with a car customization program (so, making many of the same decisions as in a deep simulator, though not to the level of granularity) -- They found that even when the number of details could not be reduced out, just the order in which you present choices had a strong impact on engagement and satisfaction of choices. The trick was building up from simpler choices to build investment and confidence (and to a lesser degree to learn how the choosing apparatus functions) and to withhold more subtle or difficult choices until later (though, no later than they would impact subsequent choices).
I think in the end the number of choices is largely irrelevant, and that what you're really chasing is that your customers are satisfied with their choices and confident of their decisions having the impact that they think they do. Another experiment they talk about is that they presented 600 magazines organized as 10 categories, and 400 magazines organized as 20 categories where the set of 600 contained exactly the set of 400 and then some -- they found that people felt that the set of 400 magazines was providing more variety than the set of 600 even though that was objectively false; reason being was that the greater number of categories allowed them to connect with what made the magazines different from the others. I'm not a cruncher myself, so I'm probably biased, but in the end I think of all these super-detailed choices with barely discernible impacts is just statistical masturbation.