As a baseline I will be using the 2024 text for Intimidation
Awe or threaten someone into doing what you want.
I like the move away from stipulating physical violence (
"and physical violence.")
Yes, and my last paragraph laid that out. GM can certainly attach a negative consequence to success -- such as the grass fire will rage out of control. But when ought they to do that? And can players describe actions that on a success hedge out such consequences?
To my reading
@Chaosmancer has laid out with great determination a dilemma. Intimidation must come with unpalatable costs (negative consequences, even on a success) and cannot be permitted to be strong enough to make those costs worthwhile (would amount to "overcoming BBEG").
One needn't agree with either horn of that dilemma, or one can blunt the first and accept the second and vice versa. If one is determined to paint oneself into a corner, one ought of course to uphold an extreme version of both horns!
I have outlined how to go about blunting one horn or other. Other methods are possible. I have aimed to illuminate the form of
@Chaosmancer's argument, and I don't find the extreme takes needed to sharpen the horns very compelling.
At the table I've observed that we're capable of letting creatures be intimidated without fostering dangerous resentment (and I find uncompelling the folksy social theories that purport to forestall that) -- thus blunting the first horn. And I've observed social skills having strong effects without swinging inevitably to breaking the game -- blunting the second.
As to groups that find themselves caught upon both horns of the dilemma through their commitment to extremes, I would first ask: why make such commitments? But if they are enjoying that style of play, then perhaps at their table it truly will be awkward to get Intimidation to work as written. Providing an example where a mechanic would appear to be at odds with a desired play style.