I admit im not educated on the rules of other editions and systems but how does a games rules ever prevent a GM from saying ‘yeah ok i think you said all the right things there I’m just going to waive making the roll/check/using whatever mechanic is meant to be used and auto succeed that one’?
Nothing, and such things
used in moderation are in fact very good and helpful. (Good Skill Challenge rules absolutely should include such things.) But the critical bit is what I italicized:
used in moderation. If you'll permit the turn of phrase, that's the exception which completes the rule.
By which I mean: No one argues that a foolish, bullheaded application of rules without context, or judgment, is somehow a wonderful and noble thing. And yet rules are still useful. How do we resolve this conflict? In exactly the same way as we explain why having tools is useful to a craft like carpentry, even though being
confined to only doing the things your tools can do would be foolish and destructive to the craft. The mere existence of a tool does not mean that it
must be used in every context. The fact that the tool is not appropriate to all contexts, and indeed may never get used in any of a variety of projects, emphatically does not mean that the tool is somehow bad, or wrong, or worthy of being gotten rid of.
Instead, well-made tools, tools that work well for what they're made to do, are a powerful aid. Instead of viewing the circular saw with suspicion and opposition because you don't
absolutely need it, the carpenter looks at such a tool for what it will make
easier to do, when it is useful to use it, and what ways it might enable new approaches she could not use before.
Rules are tools. They are made to
do something. They can be made badly, like any tool, but that simply reminds us to demand well-made rules. They are not useful in every circumstance, and even when they might be used, other tools (or even no tools at all) may be the wiser choice. Rules cannot tell you when it's wisest to use any given rule. They can help you think about it, but you still have to
do the thinking. That we can, and sometimes
should, avoid using a particular rule does not mean the rule is worthless. It means that human decision-making is still important, even when there are rules.
So the question is no longer, "why do we have this if we don't always use it?" Instead, we should ask other things:
"What does this do for us? Is that worth doing?"
"How well does this do what it was made to do?"
"How easy is it to know when to use this, and when not to use it?"
"Are there other ways we could achieve the same ends?"
etc.
And when you start asking these questions formally--with measures and testing and actual
analysis--you have begun the process of real, actual design, and not just being an auteur fiat declaring whatever you like.