I feel like these stats are kind of misleading because of their questionable usefulness. It's not like it's outright dishonest, but "performance showdown"? What are the chances that a situation like the ones they tested ever actually show up in a prod environment?
"it certainly pays to give io.js a look as out-of-the-box you get some rather impressive perf improvements." Might be true, but feels like an overstatement in relation to the data provided.