There is an argument in the roleplaying community that is as old as the hobby: to what extent do rules affect the game you are running? Do rules count, or are they just a context for fiction created by the players?
In recent years, at least, I fall on the side of "rules count quite a lot". Which isn't necessarily to say that I think you need a very large, complex, or detailed set of rules to facilitate good roleplaying.
Rules condition behaviour, both from the players and from the GM. If a game system devotes half of its page count to detailed combat rules, it is a reasonable assumption to make that combat is intended to play a significant role in this game. Does that mean that you have to play the game in line with that assumption? No, but then again, if you are playing a game system that does not support the kind of play that you are inclined toward, one might reasonably ask "Why are you playing that system?"
Case in point: the Sanity rules from Call of Cthulhu. These rules reinforce the theme that investigating the Mythos is not only dangerous to life and limb, it can make you crazy too. It increases the feeling that the characters are fragile, that the forces they face are overwhelming (based on the number of things that demand a Sanity check in a typical scenario), and that the inevitable result of play is a downward spiral. Sure, there are plenty of players who play CoC in a more guns-blazing, pulp adventure mode, but there is very little support in the rules for that. You would have to be either very, very lucky or ignore a large portion of the game system to play a CoC game that rewards that mode of play.
And if you want to play a pulp adventure style Lovecraftian game, why not run it in a game system that expects and rewards that type of play? Such as Spirit of the Century (or more recent incarnations of Fate) or Adventure! or Hollow Earth Expedition?
The default mode of play in Dungeons & Dragons (which has, it must be said, remained popular for forty years -- so who am I to disdain it?) is killing monsters and accumulating treasure. Sure, you could use it for a highly political game of wheeling and dealing in the Game of Thrones mode, but there is very little support from the actual rules for that kind of thing. You might as well be using Big Eyes, Small Mouth.
Rules are not neutral, as we claimed Back in The Day. Rules are built around expected and implied behaviour from everyone involved. You could play a game of D&D that's high on drama and character development, but there is nothing in that system to reward you for making choices that are dramatically powerful or narratively interesting, as opposed to just hunting down some hapless orcs and taking their stuff. Killing monsters and getting treasure improves your character, letting you deal with bigger and bigger challenges, creating the framework for epic adventures. This is the point of the game design. The details have changed a bit over the years, but the focus is the same.
I'm not slagging D&D. It does (and has always done) what it does very well, and that's why it remains popular (as do games like Pathfinder or Castles & Crusades which borrow heavily from the D&D playbook). What I'm implying is that choosing D&D for your fantasy game is choosing a style of fantasy game. Choosing Warhammer Fantasy or Runequest changes what your game will look like in significant ways, as does choosing Everway or Amber Diceless or Burning Wheel.
Sometimes we choose a particular ruleset because it's what our group is familiar with, and that familiarity means that the rules have a certain transparency -- we're not thinking about how to do something, according to the rules, only about what our characters want to do. That's something that can't be underestimated, and sometimes the familiar choice is a good one. Having a strong foundation for your game came make the ambitious things you've got in mind seem easier and more achievable.
Sometimes we choose a game that's new, simply because we want to give it a test drive and see how it goes. Novelty has its own pitfalls, though on the whole I am a big booster for trying new things and seeing what they can teach you about different kinds of play. There's always something that you can steal, even if a game doesn't do everything that you want it to.
Once you've got a good background in different game systems, though, I think a GM must make the choice of ruleset very carefully, with an eye toward what kind of game they want to play. You can always change rules part way through, if the game's not doing what you want to, but sometimes it's very hard to change directions once your players begin to condition their play toward what the rules emphasizes.