Game System Analysus - Rune Quest
Rune Quest (RQ) was one of the first systems to replace character classes with a freeform skill system. It was also the first system to create monsters the same way PCs were created. It also revealed one of the most in depth campaign settings (however, it was not the first RPG to come with a deep campaign setting, that honor goes to Empire of the Petal Throne, released just a year after D&D). Rune Quest also introduced the gaming world to vapor ware, claiming that the Hero Quest rules were just around the corner (rules for hero quests were finally published as a formal game in 2000 with the introduction of Hero Wars, later renamed to Hero Quest, however that game is rather incompatible with Rune Quest).
RQ introduced some new concepts like armor that absorbed damage instead of making you harder to hit. Being large made you easier to hit. RQ also had spell points (though it probably wasn't the first such system - it was the first such system I used). It also introduced character religion as a real choice and impact in play (instead of the gods of D&D that were given combat stats, but never impacted how their clerics were played).
However, one problem quickly surfaced with the game. After an adventure, characters got the opportunity to make an experience check for every skill they had succeeded in. The gamist players at MIT immediately pointed out that this perversely encouraged characters to carry a golf bag of weapons (long before D&D 3.5 introduced it's new DR system...). Characters (assuming they weren't too threatened with their life) should switch out their weapon after scoring a hit and a parry. Since everyone had an open locks skill, some groups had players try to open locks in reverse order of skill. My early solution was to count the number of successes in each skill, and add this as a bonus to the experience check (this has the problem that combat skills will get a bigger bonus). My later solution was to give each player a fixed number of experience checks (and I may have assigned some).
In my most recent RQ campaign, I noticed the only time I've ever seen real success with having different languages in the campaign. Since RQ language skill is just like any other skill, you start off really bad, and slowly improve. Of course if you play the rules strictly, there's a significant chance of native speakers misunderstanding each other (of course some would point out that's realistic - but not at the rate it would occur in RQ). There also aren't that many languages you need to learn. The result, players could learn languages, but had to make choices as to how much resource to put into it, and how much to trust using the local language as opposed to using a better understood language.
Rune Quest was also the first game I'm aware of that allowed characters to spend money on improving their skills. This creates a very good econimic flow as much of the treasure gets spent on training. Training also takes time which makes campaigns last several years of game time (and the players plot out where to spend the winter).
The 1st edition had another problem. Hit points were Constitution modified by Size and Power. Every creature in RQ gets 3d6 for CON (except maybe Dwarves, I think they got 4d6). Later the role of CON and SIZ were reversed and the game worked a bit better. There's still a problem that large creatures do more damage than most people have hit points, especially when you take hit locations into account (a decent human only needs about 8-12 points of damage to their head to be killed, giants might get a 12d6 damage bonus).
A big problem arises when RQ is examined for the possibility of replacing random attribute rolls with a point buy system. INT is way more important to skills than any other attribute (plus it restricts how many spirit spells you have available). It also is one of two attributes that can't be raised at all during play (SIZ is the other). DEX is also pretty important (though it's one of three attributes that can always be raised to racial maximum, and the only one that can be trained there - CHA and POW both increase only with experience). CHA has very little use, affecting only a small set of skills. Of course once one notices this imbalance, even rolling for attributes will be disappointing, being quite a crap shoot.
RQ is not really and ideal gamist game. It has all the trappings of a tactical combat game, but the rules are imperfect. Players can also easily fall into a trap of spending resource on non-combat skills, or joining cults that don't help the player much (some give free weapons training and/or free spells). To get the most out of the game and setting, one has to have some simulationist play also. It takes a good social contract and willing players to pull this off.
As I write this, Mongoose Publishing is working on a new edition of Rune Quest.
I have had lots of fun with one RQ campaign, and a fair bit of fun with a few other campaigns, but it seems to really be a hit or miss. If the players get into the setting, and the game drives more towards simulationism, I think it can work well. If the players simply treat it as another type of D&D and play gamist, the shortcommings of the system will blow wide open.