I am a big fan of "reverse luring" for duration behaviors. It's a variation of "doggy zen" or "it's your choice" if you're familiar with that game. RL uses treats in your hand to communicate to the dog when he's right and when he's wrong as well as teaching the dog to actively resist the distraction.
Example: teach dog to keep chin resting on the floor for longer. Once the dog has already been trained to offer the chin rest behavior, have your hand with treats held out. When your hand is outstretched and treats visible the dog is doing right, if the dog lifts up his chin to look at the cookies, close the fist but keep your hand there. Now the dog has to actively lower his head away from the fist of treats to get a reward. As the dog gets the concept this can easily be proofed by now moving your hand of treats directly above the dog's head, zoom back and forth, etc. Palm flat=correct, fist closed= dog needs to try again.
Once the concept is taught I've used it on a variety of behaviors from a solid hold of a dumbbell to proofing foot movement on stand stays. I'll see if I can find a video online, otherwise maybe I'll make one
An alternative suggestion for a nose touch duration is to require TWO noses before rewarding. Before I discovered RL I taught a "sticky target" using that method. The dogs started out doing
nudge.........nudge. click/treat
nudge....nudge. click/treat
nudge..nudge. click/treat
nudgenudge. click/treat
nuuuddgeee. click/jackpot
Basically the dogs got lazier because they knew I wanted 2 and they soon barely left the target at all before doing that second nudge and thus I had a tniy bit of duration I could reward.