D&D 5E - Reliable Talent. What the what?

My players are just on the verge of 12th level, and while I've had plenty of experience DMing earlier editions at high levels, this will be our first foray into the upper levels of 5E. As one might expect, there are a few rules that are taking us by surprise as we see them in actual play for the first time, and none more so than our rogue's Reliable Talent.

Delving into the internet, there have been plenty of threads on plenty of boards parsing this ability, but I wanted to get some fresh perspective on it if there's any to be had.

Now, to be fair, we've only played two sessions with Reliable Talent in effect, and in only one of those did it rear it's ugly head, but those instances were enough to make me want to take a serious look at it.

For nearly every skill check I threw at our rogue he simply auto-succeeded. His Perception, Sleight of Hand, Acrobatics and Stealth are already through the roof anyway, which is fine, but now, if the DC is 20 or below he may as well not even roll. And that seems insane to me.

I can't think of many (or any, off the top of my head) things that the PC's might attempt in the game that they can simply succeed at without limit or use of some kind of resource. And it's only going to get worse as he levels up. As it currently stands, I can see no reason why he couldn't tell me he's going to go out on the town on a thieving spree and my whole job would be to just hand him a list of loot as long as he isn't breaking into the local thieves guild or the kings castle.

This seems wholly at odds with cooperative story telling. With no chance of failing there is no drama. I could even see it in the players eyes when, during our last session, he listened at a door for noise and barely looked at his dice when he rolled it. Everybody at the table knew he couldn't fail, and quite frankly, it was kind of a bummer.

I've read plenty of posts online giving some pretty reasonable validations for Reliable Talent, but none of them really address the problem of the lack of drama inherent in an auto-succeed ability.

And yes, sure, I could raise DC's through the roof, but that seems cheap. As they take on tougher and tougher challenges, DC's will rise, but they need to make sense. I'm not going to counter his thieving spree with every house having DC 25-30 security measures.

I hate to nerf. The player built his character and got this far, and he should reap the rewards. And the devs built the ability for reasons that I have to believe make some sense. And I have to believe it's been play tested, and I am loath to tinker with house ruling stuff without very good reason. But this is making my brain itch.

So what I'm asking for here is how other DM's have dealt with this and how it's affected your games. Is it as broken as it appears? Have your players abused it? If so, what did you do about it?

If I do move forward with a nerf, I have a couple of ideas.

1. Implement the natural 1 rule for skills. It doesn't seem out of line that a skill attempt, like an attack, could fail on a 1. Even masters of their crafts sometimes err. In fact, I don't really understand why skill checks are exempt from the natural 1/20 rule in the first place. Furthermore, our rogue has the Luck feat, so his odds of failing a skill check would be pretty darn low. But at least there would be SOME chance of failer that would make his roll mean something.

2. Change the rule to give him advantage on skills checks with prof bonus skills. Again, this gives him a significant boost to those skills (some of which he also has expertise in) so his success rate would be very high, but again, there would be some small chance of blowing it.

ncG1vNJzZmivp6x7prrWqKmlnF6kv6h706GpnpmUqHyzscuimJuklWLBorjEp6tmr5iWwW7Ax55ksKCRqXt3fJFpaXJn