No, not as in überdesign, but as in designing more than the interaction really needs. Someone just looks silly if their text talks about putting an X in the checkbox but a user who clicks gets a checkmark. One solution is to update the text to mention the checkmark, but then some browser out there is probably going to show an X in the box. The simple solution is to keep the language general and focused on the meaning to the user of the task undertaken, something like “Select this checkbox to indicate awareness and acceptance of any and all legalese inserted above, even the fine print.” Or the user will almost certainly understand the implications of clicking if the instruction starts with “I acknowledge . . .”
The same lesson carries over in all kinds of situations. The browser will do a decent job of placing elements if we give it the minimal specs needed to relate elements to one another. The more picky we get as we strive for pixel-perfect placement, the more *other* things we have to specify in excruciating detail to work around that first element. At least until we all attain omniscience, anal retentive tinkering — with language, layout specs, and all kinds of other stuff — will feed that OCD drive to wreak our own special flavor of havoc in the world only at an ever-rising cost of time wasted and risk of unintentional silliness committed. A lesson for our times.