So as I've been working through these rule changes in my head and then trying to transcribe them into words, I inevitably get halfway through a post and ask myself "Why?" I am having a hard time giving myself a really good answer.
Some ideas would make for nice alternative rules or just as thought experiments, but when it comes down to actually improving the game, they are minor tweaks at best and just unneeded complication at worst. I am seeing this come about from my methodology of meticulously going through the rule sections. It just seems I am finding areas to change that don't really NEED it.
So in order to get myself on track with changing rules to improve the game rather than to just change rules, I am going to go with a new design methodology. I am going to start with a Design Goal and proceed towards what rules changes would best fulfill that goal. Here's an example:
Design Goal: Allow greater freedom for players to create the type of characters that they want.
This goal comes from the frustration of level limits, restrictions on class and race combination, and ultimately having people settle on what character to play. I like how 3e really opened up character generation options. While they quickly went overboard with prestige classes, splat books, and an overall emphasis on 'character builds,' I prefer to open options rather than to restrict them.
So that is why I threw out all level limits, restrictions on race/class combination, and allow any combination of classes to be used in multi-classing.
Design Goal: Giant monsters and animals are too easy to kill.
Design Goal: 1st level characters are especially fragile.
Ok, so this is post facto reasoning for the mass within hit points system. Either way, it gives 1st level characters more durability and helps to separate PCs from average people without creating "NPC classes" or "0 level" NPCs. In addition, big monsters have more hit points without hurting all the mechanics that are built around Hit Dice.
Moving forward, I am going to actually put the design goal ahead of the rule changes whereas previously I have been putting the rule changes ahead of the design goal.
The little fiddly rule changes are best made as rulings on the fly and then slowly incorporated into the rules in general rather than front loading the system before play even begins.