From the previous thread:
If there's not plausible reason for that lock to not be a run-of-the-mill generic lock, then the level 11 PC should more or less auto-succeed, assuming he/she has either Thievery training or a good dex score (and page 42 would confirm this; easy tasks are generally 5 + 1/2 level, so even if you're using level 11 "numbers" the player should still auto-succeed). And if you wanted to be even more purist, that "level 1" lock should have such a low DC (~11?) that even the clumsy fighter with his mere +4 to Thevery checks should have a decent chance of opening.
If instead the party encounters a lock installed by a group of giants (who are around level 11), then it shouldn't be a stretch to suggest that THIS lock is an appropriate challenge for the party rogue. maybe it's of giant design, or the locking mechanics are heavier/larger that common lockpicks are hard to use with it. Whatever.
Again, I'm not saying that the 4e DMG communicated this well (in fact, the DMG communicated a lot of things poorly), but as a system for improvising challenges for players, it works as intended (and I would argue, it works well).
I think this is another misinterpretation at best, and at worst, ignoring VERY good advice. Not every NPC needs a character sheet! Encouraging the DM to be constantly referencing some stats so he can accurately tell the PCs that THIS guy has a +12 to his Nature check only slows the game down. Instead, the DMG is (rightfully!) encouraging you to not make NPCs the focus of your game.Like the advice to not stat up every NPC, and to only give stats to NPCs that were supposed to engage in combat was taken to mean that NPCs don't matter unless the PCs are killing them.
And it's not like you can't create NPCs with full stats (either PC generated, or monster generated); you absolutely can! But I would question the wisdom in doing so unless (as the advice says) you were planning on using them in some sort of combat. If the NPC is important enough, write out four or five traits/quirks about him/her, and guesstimate their skills/stats if the need should arise. As much as the DM advice given in 4e had problems, it also encouraged the DMs to work "smarter" (not "harder") to deliver their game. And despite its problems, page 42 is a brilliant tool that 3.5 sorely lacked, and that I hope 5e gets.