The screen said there was capacity.
The queue said otherwise.
That was the annoying part. Nobody was technically blocked in the dramatic sense. No red failure banner. No chain halt. No obvious outage language for people to screenshot and throw into group chats. The workflow still existed. Midnight was still running. Transactions could still go through.
Just not all of them.
Not at the pace demand was asking for.
And that is a different kind of problem. Harder to explain, honestly. People understand “broken.” They understand “fees are too high.” They even understand “network congested” because crypto has trained everyone to treat pain like a visible number attached to urgency.
Midnight bends that intuition.
Because Midnight does not make execution a simple question of who is willing to pay more right now. The system separates ownership from usable execution capacity. You can hold NIGHT and still be thinking about a different constraint entirely: Midnight DUST. That is the resource the workflow actually leans on. Renewable. Consumable. Not just a token leaving your balance every time you touch the system.
Which sounds elegant right up until the usage curve gets unfriendly.
Then the pressure shows up in a less familiar place.
Demand keeps climbing. More users. More contract calls. More activity hitting the same application surface. But the execution side is not governed only by appetite. It is governed by how much Midnight DUST is available, how quickly it regenerates, where it has been designated, how much has already been consumed, and how much quietly decayed before anybody used it well.
That is the part people miss when they talk about utility too cleanly. On Midnight, willingness is not capacity. Interest is not throughput. Need is not execution. The action can be perfectly valid, the user can be perfectly ready, the application can be perfectly live, and the useful thing still tightens because the resource beneath it is finite in a very specific way.
Not money-finite.
Resource-finite.
And those are not the same pressure.
A normal fee market at least has a crude story people recognize. Demand spikes, prices rise, someone pays, someone waits. Ugly, but legible. Midnight moves the question sideways. The issue is not only what the user wants to spend. It is whether the execution resource exists in enough quantity to absorb what the workflow is asking for.
So the system does not necessarily snap.
It constricts.
That feels more native to Midnight than people admit. The protocol is trying to preserve privacy and usable computation at the same time, which means it cannot just lean on the old public-chain habit of exposing every cost as a direct, transferable fee event. Midnight DUST sits in that gap. It makes private execution usable. It also turns utility into something governed by generation and availability, not just desire.
Which gets operational fast.
A team can onboard more users and still discover that growth has outrun execution capacity. An application can feel healthy at the demand layer while the real pressure is building somewhere duller — replenishment assumptions, designation choices, usage modeling, the boring arithmetic of how much DUST the workflow actually burns when people use it like they mean it.
That is where the mood changes.
Because nobody likes opening the dashboard and realizing the problem is not adoption.
The problem is that adoption kept going after resource planning stopped making sense.
And Midnight does not solve that with a dramatic market scream every time. Sometimes the system just starts getting tighter around the edges. Things still work. Then work less comfortably. Then require more deliberate allocation. Then force somebody behind the app to answer a question users usually never see.
How much utility do we actually have?
Not in theory.
Not in token narrative.
In executable terms.
Because once demand grows independently from Midnight DUST generation, the thing deciding system support is no longer just how badly people want the workflow.
It is the rate at which private execution resource can exist, recover, and be pointed where the pressure actually is.
Which is a much less glamorous answer than “the market will handle it.”
And probably the more important one.