With the number of game system choices out there, I think really everyone can already find pretty much what they are looking for in game that they want. However, diversity is definitely a part of any gaming group, so the system has to allow each player and the GM to get what they are looking for out of the game, especially in campaign play.
Understanding how systems handle ranged attacks or spells in melee combat can be critically important when you have a character that simply can afford to allow enemies extra attacks on them. Plus, all of this simply is just a modeling of the fantasy world with mechanics and dice added for chance, so the more you understand the mechanics you can better "play" your hero. There's hope that a GM will rule consistently and fairly, but we've all seen cases where that is simply not the case, or for sure might not seem so at the time.
In the HitDice system I really tried to keep it, at its core a "fast conflict resolution system". I do like the "Epic Battle" idea, but don't feel the need for every encounter to be played out in that scale where it takes two hours to resolve a fight with a dozen goblins.
To do this modeling, I essentially started with a skirmish system, one that I was designing for a board game environment, that was designed to be fast paced. As I began modeling the combat, I liked how quickly they came to a resolution. After a number of play testing scenarios, I realized that even with these simple mechanics, there were a lot of "levers" that could be pulled to model different situations, and they really could grow in complexity.
In general HitDice is meant to handle the "old-school" adventures with 30+ rooms (maybe not all of which will be thoroughly dealt with), along with quest hooks, the journey and boss battle in a session or two. So you can imagine that the resolution of any particular task is handled relatively quickly. Thinking of the earlier RPGs, most encounters fell into the "non-critical" (or filler) category, designed more to paint the environment than something critical to the story. So why should these take longer than a few minutes to resolve?
Skills are normally fast resolution, but I felt they should land more on a probability curve than a linear one, and everyone should have some chance to do anything (some things still require training, such as spell casting)...though this chance may have odds of 1 in 1,000. With linear odds the result is more about who rolls higher on the dice, where end up with things commonly occurring that just don't make sense, the 18 Str warrior fails to open a door, and the 10 Str Mage comes up and rolls a "Nat 20" and succeeds.
I also agree many systems do this well, but I think dice pools do a much better job of capturing results than system that use linear-systems/one-dice (d20 or d%). On top of the pool. the exploding dice add essentially an unlimited degree of success that can really generate 1 in a million probabilities, which at critical times can generate extremely heroic results, as well extremely unfavorable results when achieved by the opponents.
I know HitDice is not the only one that allows for this, but what I've tried to do is to keep the mechanics simple. So that they work even in simple board games very similar as they do in more complex RPGs...one system to govern all scenarios. Other similar systems have much more complex methods of determining success and/or the number of dice requires special utensils and a large surface space to allow the roll to be made.
The system has been designed as an evolutionary mechanic that one can learn in one of many board game settings, and then progress into and overlord/referee system and ultimately into a full blown RPG environment with little change to the game mechanics at their core, but simply by expanding the player options and integration of the story telling aspect into the game.