You have to understand that some of us find this extremely unsatisfying. It's putting the description of the game world secondary to the metagame mechanics. The setting conforms to the rules, not the rules conforming to the setting.
I can somewhat understand it, but I think a good part of the problem may be an issue with how 4E presents it.
When 4E talks about 'scaling to level', it makes it sound like it is saying, "Here are the DCs to scale to the levels of the PCs."
What it is
actually saying is, "Here are the DCs to scale to the level of the
challenge."
Now, most times, those will be around the same thing, assuming PCs are facing level-appropriate obstacles. But it means that if you want the 5th level party to run into a level 15 lock, you know the DC for it. Or if the epic level rogue runs into a level 5 lock, you can confirm that, yes, he can open it up without even needing to roll.
At its heart, that really isn't any different then what we had before. In 3rd Edition, it tells us we have four types of locks: Simple, Average, Good, and Superior, with scaling DCs from the easiest to the hardest. In 4E, we have a broader scale of DCs, abstracted a bit to represent not just the lock itself but also the circumstances around it, thus justifying the broader range of DCs.
Now, all that said, yeah, we are still deciding elements of the game based on how challenging we want the lock to be, rather than other elements. But aren't those other elements arbitrary anyway?
I mean, why is a Superior lock DC 40 while a Simple lock is DC 20? Is there an actual in-game explanation for it, other than, "One lock is better than the other". An in the end, the DM is the one who chooses to place the lock in a scene - isn't he going to be doing so based on what is appropriate to the scene? The PCs try to break into a commoner's house, he'll probably have a Simple lock. They try and break into the house of a powerful noble manipulating the kingdm, and he has a Superior lock. Try and break into the ancient wizard's tower, and he has a Superior lock reinforced by an Arcane Lock and other wards.
Is that any different from declaring that the commoner has a level 1 lock, the noble has a level 10 lock, and the wizard has a level 20 lock? Or whatever else the DM feels is appropriate for that figure?
Honestly, I find that more robust way of being able to determine details. Because if we actually look at the 3rd Edition locks, probably the biggest "in game mechanic" to them is their price. And the best lock is, what, 150gp? That's... relatively cheap, in terms of the money these games throw around. Which means that past the first few levels, shouldn't most every lock the PCs run into be a Superior lock?
It would me the most logical conclusion supported by the description of the game world, perhaps. But most DMs probably wouldn't do that, and be more likely to use them sparingly, even though the price difference between the Simple lock and the Superior lock is so trivial. Again, they will be making decisions based on what is appropriate for whomever owns the lock and based on its purpose in the game - an obstacle for the PCs to overcome, a bit of flavor-dressing, a barrier to something out of reach.
And for that purpose, I find having scaling DCs available at hand to be a very good tool to have.
Now, all that said, I find they are most useful when presented alongside some more set in stone DCs. I want to know both what DC would represent a challenging jump at level 20, but I also want to know what DC lets a PC jump over a 10' pit.
But I don't think that the abstraction of the scaling DCs, itself, artificially puts the description of the game world second. It remains a tool used by the DM to capture the description he has already come up with, just like any other method he might use to arrive at those DCs.