I disagree completely with this. If you want to truly understand something you need abstractions in order to strip away the non-generalizable bits from concrete examples. In physics you have particles, in chemistry molecules, in biology cells, and so on. Obviously not every particle, molecule or cell is the same. But treating them such allows the discovery of things applicable to all (or at least most) cases of each class. And even the exceptions are interesting for the sake of being exceptions.
“Content” is a similarly useful abstraction, and if anything, we need to be reminded of it all the more these days. When the Web was started, it was understood that we needed to separate the actual content from presentation, and that the user should be the one to decide what fonts to use (if indeed any – the content could be spoken for blind people or used as input to a program, etc.). This useful feature is being lost by people who can’t abstract the content as being different from the presentation.
