I'll ask it again: why is the idea of running Safeguard not even taken in consideration?
Because Safeguard only lasts 5 turns and has to be used before Darkrai starts sleeping things. (I don't even know what the distribution is but that could play a part as well) All that accomplishes is continuing the trend where you have to play extremely aggressively and quickly in order to win games. (or sling Lum Berrys and Sleep Talks everywhere in the hopes the opposing Darkrai player is stupid enough to not scout a set first) I've actually played somebody who used Safeguard Wobbuffettand got lucky enough to dodge my Sleep Powders to Encore my Subs and set it up. The problem is that it has a limited amount of turns and only addresses sleep. If my Flutterby can stall out the time limit and then continue to shuffle things around a Darkrai would have no problem just shooting out some Dark Pulses or pulling out to a teammate. (which the ladder seems to have a lot of difficulty understanding)
Are you guys really having trouble with Darkrai / sleep? It only took me a few matches to get the hang of this new metagame. Even more, it only took me one or two minimal changes in my pre-existing team to make sleep a non-issue. It isn't hard for me to minimize the number of Pokemon asleep to one. The increase in Darkrai usage is quite exploitable, especially the Sub NP variants - I don't see what's so shitty about this metagame tbh. As legitimate stall users have posted, Stall isn't screwed by sleep either if they choose wisely their sleep absorber.
Of course if you expect to use the same stall team you used half a year ago or make little effort in dealing with Darkrai, you're gonna get raped. It's just a different metagame that you have to learn. However, sleep itself has been manageable so far without the need to clause it, from what I have found.
The thing is, basing your conclusion on ladder experiences alone is unreliable as the Ubers ladder is notorious for having very poor players. (which heavily outweighs the few who know what they are doing) I haven't seen any of them that have learned the value of keeping their Darkrai safe and those who may have caught on have such generally crappy teams that they crumble before they could even use it.
I'm interested in how you can exploit sleep outside of mono sleep talk genies. Unless you are talking about surprising an open Darkrai with a random Sleep Talk/Lum Berry which shame on them for taking unnecessary risks to wrap a game up early. (which what I have a strong feeling is going on with these stall teams but I haven't ghosted them so they'll need to be a bit more exact with their experiences)
Sadly, running Lum Berry/Sleep Talk is not only taxing your walls even further just to deal with status but they aren't even that effective as Fireburn pointed out. Lum Berry mons are just one time checks that can't even switch in unless they predict a Dark Pulse/NP and he isn't hiding behind a Sub. (btw, SubRai isn't the only good one just the easiest) Sleep Talk is random not only for move selection but also for the sleep counter as clicking it when you wake up isn't doing you many favors. On top of that, while you are asleep you can not rest and you take extra damage from Darkrai's Nightmare which means, on top of hazards and 135 DPulses, they won't be around for very long. Hydration/Vital Spirit/Insomnia mons are all very poor mons in general which means they are easy to exploit by the rest of the team and usually don't even have the staying power to last long against a smart Darkrai, anyways. (come on, Flutterby can beat Manaphy) Moreover, unless you run some form of anti-sleep measure on every member of your team (which'll make it easy for a the other 5 members are Darkrai's team to pick you apart) you are going to have to make predictable plays that are going to be easily exploited as a single mispredict when dealing with sleep could mean losing an additional member of your team. So to sum it up, no sleep clause is putting a lot of pressure not only on teambuilding but also on playing and, even then, the counter measures can not last into the long term. (Darkrai isn't just a sleeper you know)
Basically, this metagame is a Deo a/s metagame where you try to setup your hazards with mons that outspeed Darkrai and then follow up with multiple attackers to keep the game in the short term and deny Darkrai as much play as possible. Keeping in my doug's characteristics of a desirable metagame, this isn't one and sleep clause shouldn't be removed.