I didn't believe that was an actual quote from the article until I read it. It sounds like a joke and I can't imagine anyone reasonable would say that.

There are definitely good reasons to build your own framework but performant and stable applications isn't one of them. Usually it's a solution looking for a problem.

At Google or Facebook that makes sense. If you encounter much larger scale than pretty much any other company, it's hard to find libraries that work for you. So the observation was correct but they should've concluded that this is just a result of scale and complexity, not a general strategy.

Keep in mind, Asana was founded in 2008.

Luna had some of the same concepts underpinning React, about 5 years earlier.

That was in 2009 apparently. Maybe it made sense then.

just use nodejs

weird. Not sure how they would have been hired at Google or FB with this mindset.