My first thought reading this is, well, "conversation doesn't scale very well", as David Weinberger said (regards some recent Reddit contretemps).
Looking at NPR's own article, I'm findng the justifications given to be strongly suspect. NPR's focus on how many participants are engaging, and from what platforms, rather misses the boat. It's not that I feel the decision itself isn't without merits, but the merits given are exceedingly poor ones. I'm hoping they're not the ones actually used.
I've done my own measurement of public user activity on large sites.[1] In the case of Google+, my and Stone Temple Consulting's independent analysis[2] showed that 0.3% of all users actually engage in public posting on the site.
The real question isn't "how many users are commenting", but "what is the quality of the comments received?". Following critiques of my first study, I performed a second follow-up looking at where intelligent conversation was happening online, using the Foreign Policy Global 100 Thinkers list as a proxy for intelligent conversation, and the arbitrarily selected string "Kim Kardashina", as its obverse. This gave the world the infamous FP:KK index -- the ratio of mentions of any of the FP Global 100 Thinkers per instance of a spin-off of the OJ Simpson trial fall-out.[2] A few results stood out -- Facebook's scale, Reddit's relatively high quality, Metafilter's phenomenally high S/N ratio, and the amount of high-quality material posted to blogs (though perhaps never seeing the light of day). I'd be interested in some further follow-up along these lines.
One of the problems is that at Internet scale, Sturgeon's Law is far worse than six-sigma compliant.[3] Not only is there an awful lot of crud, but simple mechanics mean that no one person can see more than the tiniest fraction of what is transacted online daily. The simple acquisition cost and assessment of "is this worth reading or not" is absolutely prohibitive.
Or as Clay Shirky says, what we've got isn't content overload but filter failure.[4]
On the problem of idiots, my personal solution is simple and surprisingly effective: block fuckwits.[5] At scale, this evolves to the problem of figuring out who is and isn't a fuckwit. The ability for individual actions against specific authors and publishers to be applied generally strikes me as a useful tool. Not a complete fix (there are controversial voices who do deserve to be heard). But if the cost of being an asshat is being an asshat screaming into the void, part of the problem is addressed.
(Yes, this means some form of 1) persistent reputation, 2) tools for applying reputation as filters, and 3) limitations on newly-created profiles. The idea of vouchers (again, with a reputation penalty applying) to bootstrap new identities may help. There's much in common to approaches against email spam in this.)
NPR in particular are counting the fact that many messages come from desktop users as a failing of their system. I see that as absolute insanity. Desktop systems are hugely more useful than mobile devices for composing content. Especially thoughtful content. I know this because I've been trying, and losing, that battle myself, using a 9" Android tablet and Bluetooth keyboard -- one of the better mobile authoring configurations possible, and it still stinks. The six lines by 45 characters I can see in HN's edit box certainly don't help. I've written a long rant at Reddit on this specific problem.[6] Whilst composing this comment I've had Firefox/Android crap out from under me, continuously popped out of the edit box, and unintentionally navigated from the page. Thankfully persistence of user state in edit dialogs has improved slightly, but the experience is frustrating to say the least.
I'd use a proper editing environment, say, vim, except that under Android, VimTouch doens't interact with the clipboard. I can neither copy content into* it, nor out. Termux's vim client is slightly better -- I can paste through the Termux clipboard, but copying out is virtually impossible, and doesn't capture more than one screen at a time.
The fact that few people are entering thoughtful comments on mobile likely says far more about the state of mobile tech than it does about NPR's audience.
________________________________
Notes:
1. https://ello.co/dredmorbius/post/naya9wqdemiovuvwvoyquq
2. https://www.reddit.com/r/dredmorbius/comments/3hp41w/trackin...
3. https://www.reddit.com/r/dredmorbius/comments/1yzvh3/refutat...
4. http://www.cnet.com/news/shirky-problem-is-filter-failure-no...
5. https://plus.google.com/104092656004159577193/posts/drLZV8sm...
6. https://www.reddit.com/r/ideasfortheadmins/comments/4y2k5r/l...