Have you actually experimented much with various options for holding a camera steady while composing on the LCD? Your original post described such a severely sub-optimal "snap shooter" approach that it seemed you had not gone beyond looking at what snap-shooters often do. If you have tried other methods besides the "unbraced" approach, why only mention it in your complaint, instead of addressing other more sensible approaches to steady camera holding?
I have experimented.
I used a camera with both EVF and LCD views while shooting (Panasonic FZ5) and shot a target while (as closely as possible) maintaining constant camera/target distance. IS off for all shots.
I shot ten shots with the camera braced against my fact while using the EVF and ten shots holding the camera at a comfortable distance from my face and shooting from the LCD. (5 EVF, 5 LCD, 5 EVF, ....)
There was no appreciable difference between the sharpness/fuzziness of the images. IMO.
I'm reasonably convinced by these results (since I started with a 'to the face'/EVF bias).
Two other people on DPR ran roughly the same test and obtained the same results.
A better study (more subjects, more trials, more objective measurement system) might make for an interesting undergrad research project.