To me, this is the point where I feel it's more of a boardgame than a roleplaying game (even if technically not true). Fiction first, I say: If it just doesn't make any sense in the fiction, then make a ruling to override the mechanics. And if it's an ooze-centric campaign, make a new rule.
I agree entirely. My approach to mechanics is that they are an aid to adjudication of actions that apply to most cases. They are useful as a guideline to help DMs adjudicate consistently in most situations, and to help players assess their odds of success at most actions. However, if applying the mechanics leads to an absurd result, the mechanics have to be overruled.
I see systems and DMs that try to apply rules to all eventualities making two common mistakes (IMO):
1. Force the general mechanic to be applied in all circumstances. This leads to absurd results and makes the mechanic more important than the narrative. The 4e rule for knocking an opponent prone is an example of this.
2. Trying to make a rule for every possible eventuality, or at least as many as you can think of. This leads to a proliferation of rules that must be remembered even though they are rarely required, like aspects of AD&D initiative. It can also lead to a lot of work for the designer, like having to assign, for each weapon, a reach and speed factor, not to mention "to hit" adjustments for every armor type; which in turn increases the probability that users will find at least some of the numbers to be wrong or counterintuitive. It doesn't actually cover every eventuality, which means you haven't actually avoided the need for rulings outside of the rules.
And the more granular the rules, the more likely your armchair physics will be wrong, or at least debatable. Take weapon speed factors: it is at least debatable that reach is a more important factor than the agility of the weapon in determining who strikes first, even when you aren't charging. At best they cancel each other out to some degree. The issue is avoidable with AD&D speed factor, since those who
recognize its patent idiocy disagree with it can excise it from the game, but I would suggest that picking a side in what is, as written, a non-optional rule leads to an lot of unnecessary arguing (like in this forum).
Moreover, I would argue that both forcing a general mechanic to be used, and trying to write rules for every situation, stifle creative play, because they box players into the rules. If you want players to try to improve their odds of survival through intelligent play, you need to signal that they are free to try things that the rules don't cover, and that you won't just shoehorn every action into the existing ruleset. For example, if I am parleying with a thief, and I have my sword pointed at his chest, which he must get past in order to stab me with his dagger, and he
still gets the benefit of speed factor, its going to discourage me from trying to engage with the game world narratively. (Also, its going to piss me off. True story.)
A less controversial example might be that clever narrative investigation should be able to obviate the need for a thief to check for traps. If there is no way to find traps short of a "find traps" roll, you can't expect your players to attempt narrative ways to find a trap.