Nick D’Alosio deserves a ton of credit. He's smart, creative, young, ambitious, pulled together a good team and created a great looking app but I don't believe this will solve the problem.
Before I created Skim That[1], a human powered news summarizing website, I looked into how Summly and other TL;DR algorithms worked to decide which direction I wanted to go in. I concluded that, even if you create the best algorithm that takes the best sentences from a story, you're still going to end up with a poorly written and out of context paragraph.
Try it yourself. Read a story and try to manually find a few sentences that best summarize it. Then, without rewriting anything, try to put those sentences into a paragraph or two. You'll find that you end up with a confusing summary that's difficult to read.
The amount of words Summly limited itself to (one iPhone screen worth) is brilliant. It keeps things simple and easy to skim through but what's produced by the service is so short that it should be called a "tease" not a "summary".
Of course, the Summly app does hit the nail on the head with a summary every once in a while but this hit or miss style makes for a bad experience.
Summarizing algorithms aren't the automation challenge, those will never work right as demonstrated by the "try it yourself" challenge above. The real automation challenge is rewriting the important sentences in such a way that they make sense together and provide proper context.
On Skim That, I'm trying to solve the problem by crowd sourcing the summarizing process and asking readers to summarize a story. I've gotten a few people to give it a shot and several of those continue to summarize stories every once in a while. My challenge is how to get people more interested in participating and how to get search engine traffic to a website that goes against many SEO principals.
Edit: Human or machine, I think the summary race is great for readers and helps people be more informed. I wish Nick a lot of luck and hope that one of us solves this problem soon.