I have mixed feelings about how traps are portrayed in the latest edition.
Traps did have some problems in earlier editions. I played far too many RPGA modules where if you didn't have a rogue you were shut-down because there was NO way to bypass the trap without taking a crazy amount of damage. So keeping that problem in mind, 4e's traps are a great improvement as they offer experience and always have alternate ways of disarming them or bypassing them.
I also enjoy the idea of including traps in combat as a deadly form of terrain or a hindering mechanic. It adds a lovely dash of spice to mundane combat.
The DMG2 has some great advice for positioning and planning traps (on page 65) which I would recommend every DM read. I disagree on point #6, which I'll discuss further down.
That said, there are some things that just don't work well with traps.
The first problem is the skill DC problem. Trap DCs are just too low. This is the inherent problem with the skill DCs, or rather the lowered ones. I agree with the reasoning behind the change, as the originals were too high for skill challenges; the high numbers took untrained and unskilled people out of the equation for skill challenges.
However, unskilled and untrained people are seldom going to be the people disarming traps. There are three classes that can be trained in Thievery: artificers, rogues, and warlocks. The new DCs greatly benefit the first and the last because neither is likely to have a high Dexterity: artificers have very high Intelligence while Intelligence is also the warlock kicker stat so Dexterity becomes a dump stat, thus both classes - if trained in Thievery- will likely be going at traps with just the training bonus. The default numbers are passable for this.
Rogues are different. Their primary stat is Dexterity and they will always have an 18 or 20 in it (unless the player is going for a personalized build in which case it might only be a 16, but this is less likely). Starting with a +9 means that even on a "1" they hit a DC 10. It becomes very unlikely for a rogue to fail to disarm a trap, let alone fail by 5 or more (at 2nd level they can only fail a Hard DC trap by 1-4).
To challenge my halfling rogue I need to add higher level traps (generally 2-6 levels higher than the party) but that also greatly increases the damage output of the trap and the experience. It seems wrong to arbitrarily increase a trap’s DC just because a character is good at what they do and synergizes well; I wouldn’t increase a monster’s AC just because the fighter had the best sword and made good feat selections.
For a game that is all about specialization, it seems odd to allow such a big part of the game to be so easily overcome without even trying. Increasing the DCs for traps might penalize parties that rely on an artificer or warlock for trap disabling, but that’s why there’s the Skill Focus feat. And I can easily see a feat that allows high-Int characters to use their Intelligence for Thievery instead of Dexterity.
But that's okay, because disarming traps has suddenly become a skill challenge, so they have multiple opportunities to roll that “1”. Ugh. This sometimes works but other times having to spend four rounds disarming a single trap is tedious and much less interesting that using powers to interact with combat. It's a little like being a 3e cleric or fighter, with no option for action. "For my standard I guess I make my Thievery check. DC... 28. Then I just stand there I guess." Except for grinding-long fights it’s often easier to ignore the trap and just plow through the combat and move out of the room.
I also don't feel 4e traps work well with classic adventure (classic being every adventure with a trap published between 1974 and 2004). Most of those traps violate the DMG2 trap point #6: Traps in Unexpected Places, which argues that while it might be "realistic" and effective to put a trap in a surprising place it slows down gameplay leading to players testing every hallway.
It seems like a shot against ever Gygax module every written, where there are continual surprise and gotcha! traps (I also take umbrage with the "realistic" in quotes because it's dismissive of people wanting to inject realism into their games). I agree that having traps in every hallway (or even every tenth hallway) might be overkill and will slow down the game, but sometimes a nasty surprise keeps players on their toes. And there's nothing wrong with slowing the pace of the game to emphasise the exploration aspects of the game. Removing traps from the hallways seems to be an attempt to speed-up players to the room where they can get onto the next combat encounter. Hallways are more than just places you travel to get to the next fight, sometimes getting there is half the fun.
The DMG2 subtly introduces traps that address my middle point (the skill challenge traps) but which also work with my complaint above: minion traps. They're not described and you're not given any advice on creating them, but there is one in the examples.
I introduced minion traps into my game a few months before DMG2 and found they worked well and captured the feel of classic traps. If I were updating older modules such as Tome of Horrors or Temple of Elemental Evil I would rely heavily on minion traps. They slow the PCs down and make them think before just throwing open every door in the ancient dungeon, and they only take the rogue a single round to disarm. And while they do damage it’s generally not enough to do more than sting, but adds-up over repeated rounds or multiple small traps.
