I take FallenRX to be saying the following:
* In the case of the BitD clocks, the number of segments in the clock corresponds (in some rough sense) to the difficulty of the task/race/contest at hand;
* In the 4e skill challenge, the complexity of the challenge doesn't correlate to some "objective" feature of the task at hand, but rather is a purely metagame decision about how long the scene will last for.
This is a core feature of 4e - the minion vs standard creature distinction is just the same, namely, at its core a metagame decision about how long the fight with the creature should last.
You also see it in BW (in some contexts - eg melee combat, social influence) and HeroWars/Quest - the choice to resolve a conflict via a single roll or an extended contest.
It is present in Marvel Heroic RP/Cortex+ Heroic in a more subtle fashion: the GM can extend a conflict by using Doom Pool dice to buff a NPC opponent. Of course there should be some fiction narrated to explain what the Doom die correlates to - but likewise in a skill challenge, the GM will be narrating fiction that explains why the conflict is still alive.
In the MHRP case the GM is limited by a budget - the Doom Pool is finite. In 4e the GM is not budget-limited in that way, but the more complex the GM chooses to make conflicts (via higher complexity skill challenges, or via standard as opposed to minion creatures) then the more XP the players earn - which bring treasure and levels - and the more rapidly they meet the requirements for earning action points. So the GM's "unlimited" budget feeds player-side resource development/renewal.
Threat clocks in AW are, I think, also somewhat similar to 4e in this respect - ie serving a metagame purpose rather than modelling an "objective" difficulty. A threat clock in AW progresses based primarily on the GM's narration of the fiction -
moves - in response to the player action declarations as mediated (where applicable) via player-side moves. The GM can choose to use their narration to bring a conflict to a climax, or to draw it out.
Needless to say, I don't think that this technique, across these various implementations, is "absolutely terrible".