I don't believe you. And unless you are secretly a google programmer, there's no way for you to know if that is truly the cause of one web page ranking higher than another. Search ranking is influenced by so many factors.
Lets think about this logically. The internet is old, HTML has evolved significantly since the internet's infancy. What was perfectly written HTML code 15 years ago no longer is, yet the information presented is still valid. The goal of a search engine is to present the searcher with the most informative "answer" to their internet search. It is not in the search engine's best interest to put any weight on code validation whatsoever.
As far as spiders getting hung up on poorly written code, I didn't say it's ok to write poor code, I said it doesn't matter if it's valid.
Yes it matters if you don't have clearly defined hierarchal content, your code is junked up with tons of Java, you don't have any <A> tag links, but it doesn't matter if it's tableless or xhtml compliant.
It doesn't matter at all! It has 0% influence on
SEO!