In brief, the combat system simulates a world like players of the classic gamebooks expect, and, importantly, the mechanics are quick. Deceptively quick, given it uses a percentile system where you might have expected a simpler d10 system. The key thing is Gary Chalk's multiple combat system. Follow this:

The rulebook on page 14 describes a lone Human bowman trying to fire at a Giak and, after firing modifiers, having only a 7% chance to hit. Modifiers seem to be tiny: the Play Sheet in the sample scenario says the difference from armor was only -2%. But then get 10 archers in a group and 7% becomes 70%; go to short range and it becomes 140%; and the effect of armor is an important 20%. You roll ONCE and all 10 archers have inflicted their casualties for the turn.

Now follow this. 4 Human archers each have a 19% chance to hit, and 6 Giaks each have 14%. Add and compare, and the Humans are at a slight disadvantage: the game will come down to strategy, cover, and all the other fun stuff you expect. Now change the granularity to a d10 system. How? Does everyone get an individual 1/10 chance? Then Giaks win by superior numbers. Giaks get 1/10 and Humans 2/10? Then Humans win by overpowering the enemy. Either armor, cover, and so on all have to become meaningless, or you have to create some alternate system that involves more (and slower) die rolls.