RPG Design: Class-based vs. Skill-based, Parts Two and Three

Rampant Games' Jay Barnson continues his analysis of skill- and class-based systems with two new RPG design blog entries (here and here) that go into much greater detail about the strengths of each than the first entry did. His perceived strengths of a class system:
Easier to Design Around: In CRPGs (and pen-and-paper (modules) not customized around specific players and characters), the same abstraction of roles that makes the game easier for new players to get around is just as valuable for designers and developers. Content, AI behaviors, NPC dialogs these can all be easier when you can create it for a limited palette of character types.

Class Explosion and Inflation: This is more of a problem with pen-and-paper RPGs when publishers are trying to entice players to buy more books, but also relates to CRPGs based on these systems, or CRPG expansions / DLC. Adding a new class that fits a particular niche or blends two other classes together becomes an easy (and brain-dead) way to add player content to the game. This became particularly bad in the D&D 3.x era, when almost every book featured new core or prestige classes that were often nothing more than a variant combination of game mechanics with some kind of tacked-on class description. Since publishers (and developers) really want to sell expansions, there was always a temptation to inflate the power level of these classes so that they were slightly more powerful than the original classes. After all, who'd bother playing a new class (or buying an expansion that activates that class) that is weaker than the originals, right?

And where he thinks a skill-based system's strengths lie:
Gradual Progression: This may or may not be a feature of your skill-based game. While level-based progression is normally paired with class-based systems, that's not necessarily the case. But more often than not, skill-based games offer the players the chance to build up their skills or attributes gradually rather than massive increases all at once. In a system like The Elder Scrolls or the Call of Cthulhu dice-and-paper RPG, this could be a chance at an increase based on usage. In a game like the Hero System or World of Darkness, you can purchase your increases directly with experience points (or some similar analog). Gradual progression is more realistic, but lacks the inherent milestones of level-based progress, which is a great reward / gratification mechanic.

Skill Combination Imbalance: It's a lot harder to balance skill-based games, and one of the challenges comes from the difficulty of testing all possible skill combinations. If the game system is (interesting,) meaning that there are several skills that can interact with each other in a course of action, there's a chance that some (combos) of skills are far more effective than others. If it's extreme enough, this can cause game balance issues. One example of this is in Dungeons & Dragons 3.0, where the reworked (Haste) spell which was intended to make higher-level melee characters moderately more deadly in combat - could be used to allow spellcasters to double-cast spells every round, which was an extremely powerful side-effect (and corrected in 3.5). And speaking from personal experience, the challenges of balancing feats like Dual Wield, Speedy, Auto-Fade and the various special attacks in Frayed Knights felt like it was constantly fraught with peril. I still don't know if I got .m all right, but at least they don't seem terribly broken.