If all I'm doing is re-coding a well known algorithm in the latest fad language then maybe I buy it.
But some organizations (like my employer) are pushing the bounds of what computers can do. Either through original research or development based on academic research.
I heard an interesting talk today about distributed "eventually consistent" caches which are kept up to date using epidemiologic (also known as "gossip") algorithms. Its a slick kind of idea that allow an overall reduction in cost by letting us do more with less hardware (this year's theme).
The organization's success depends on lowering costs like this and creating new capabilities like this is most definitely an "accomplishment".