I was aware of those errors, they may stop the code from validating, but they haven't caused any issues with my target browsers. (pretty much all browsers except the dinosaurs and the more off the wall kind)
I almost spent the time getting rid of the body tag margins, background images in table cells, and name attribute for images, but (don't shoot me here) since they display properly in all browsers that I tested on, I saw no reason to crack down and remove or work around them other than to get that cool little icon that people can click on to see that it is valid HTML.
Don't get me wrong, I don't consider myself a sloppy coder, but if it works it works. Plus, I have seen sites which do validate and do cause display problems in some browsers, so although the validator is a great tool to debug HTML, I no longer swear by it like I used to.
Also, I am a fan of all CSS and div designs, but sometimes a tabular layout is easier IMO. I know the history, tables are for tabular data and divs are for display, but tables have been used by designers for years to display more graphical pages, and I personally don't see any real harm in it. The code isn't overly verbose in my pages, it is actually structured really well from an SEO standpoint.
Anyhoo, I still have more to learn so I'm open, just wanted to play devil's advocate for a moment.