I'd agree with the sentiment, but for a different reason. I think the method can be useful and maintainable, but, the problem with this implementation seems to be a mix of design (font colors, sizes) and functionality, and a lack of documentation.
Splitting the CSS into two distinct pieces (or better: files, but I realize that's a limitation of jsfiddle) would go a very long way to making this technique maintainable. Having documentation that then shows the required HTML structure and the necessary classes to add would help immensely.
Essentially, build it like a UI library that just happens to be written in CSS.
Right now, it's just a demo of a technique that's not built in a reusable way, and you need to be pretty decent with CSS in order to understand and use it (since you pretty much have to re-implement it every time you want to use it). To be fair, the HN submission and site don't make any claims it's anything but this.
Splitting the CSS into two distinct pieces (or better: files, but I realize that's a limitation of jsfiddle) would go a very long way to making this technique maintainable. Having documentation that then shows the required HTML structure and the necessary classes to add would help immensely.
Essentially, build it like a UI library that just happens to be written in CSS.
Right now, it's just a demo of a technique that's not built in a reusable way, and you need to be pretty decent with CSS in order to understand and use it (since you pretty much have to re-implement it every time you want to use it). To be fair, the HN submission and site don't make any claims it's anything but this.