QR Codes tend to polarise opinion - the cool kids think they're a really neat idea, and the REALLY cool kids make a great show of telling everyone how bored they are with QR Codes already. I think an approach somewhere in between the extremes is warranted here: explore them, see how they might work for your library, and most importantly see if your particular patrons will enjoy using them.
As I said in this presentation on marketing with new tech, what matters with all this stuff is market penetration - in other words, it's not how good / bad something is, it's how much our users engage with it. Whether you, as an information professional, love or hate QR codes is really neither here nor there - if the stats say your patrons like 'em, you should be using them, and if the stats say your users are indifferent to 'em, then park the whole idea, stop flogging the dead horse, and come back in a year's time to see if the market penetration is any deeper.
A recent article in Mashable suggests that QR Codes' days are numbered, which is fine. We can use them for now and then ditch them later, as long as we move with our patrons. The article suggests that Mobile Visual Search (MVS) is much better - rather than scanning a code like with QR, an image or logo itself becomes the trigger for your phone performing an action (to take a boring example, going to a website - though it's worth keeping in mind that QR can do a lot more interesting things than that; see the new tech presentation for examples).
In the book I mention Blippar (an example of MVS - it's like augmented-reality QR Codes on steroids) as an example of something which is much better than QR Codes, but which I wouldn't recommend investing time and money in for libraries. Here's what Blippar does:
It's amazing isn't it! Doesn't matter, though, right now, because the majority of your users won't have the tech and the apps and the knowledge to want to engage with it, so its amazingness isn't really the key thing from a marketing point of view. Market penetration is key, not the tool itself.
There's one more reason why it doesn't matter if QR Codes die soon or become a permanent fixture in our marketing toolkits: if they ARE superseded by something like Blippar, much of what we learn experimenting with QR we can then apply to Blippar, as Stephen Abram acknowledges at the end of this post. The concepts, ideas, and methodologies will be much the same, even if the tool itself changes.
So: don't ignore QR codes, and don't base an entire marketing strategy around them either. Try them out, keep statistics, make meaningful comparisons, and let your users guide you.