I've never been a strong proponent of visual validation (the practice of using code to check if something that you are looking at on screen is what should be displayed on screen). It’s something I have tried in the past and it never really seemed to live up to the hype. The idea of being able to take a screenshot of a site and then compare it to a known good copy sounds great, but the reality was always underwhelming.
In the past, I have diligently written code to take screenshots and then do a pixel-by-pixel comparison to see if they match. The problem was they very rarely did: variables such as different machines running at different resolutions came into play, slightly different rendering engines could result in color changes that were invisible to the human eye, but absolutely different to a computer. You then have to worry...