Kottke and others on standards and semanticsk
Kottke provides a good overview of Web standards and semantically correct site development. Jason points out, as many have, that just because a site validates to the W3C does not mean that it is semantically correct. Actually there are those that take umbrage with the use of the term semantically for (X)HTML, when many consider it structural tagging of the content instead, but I digress. A "valid" site could use a div tag where it should not have, for example where it should have been a paragraph tag instead. Proper structural markup is just important as valid markup. The two are not mutually exclusive, in fact they are very good partners.
One means to marking-up a page is to begin with NO tags on the page in a text editor then markup the content items based on what type of content they are. A paragraph gets a "p" tag, tabular data is placed in a table, a long quote is put in a "blockquote" tag, an ordered list gets "ol" tags surrounding them with items in the list getting wrapped with "li" tags, and so forth. Using list tags to indent content can be avoided in using this method. Once the structure has been properly added to the document it is time to work with the CSS to add presentation flair. This is not rocket science and the benefits are very helpful in transitioning the content to handheld devices and other uses. The information can more easily scraped for automated purposes too if needed.
It is unfortunate that many manufacturers of information tools do not follow this framework when transforming information in to HTML from their proprietary mirth. A MS Word document creates horrible garbage that is both non-structural and not valid. The Web is a wonderful means to share content, but mangled markup and no structure can render information inconsistent at best, if not useless.
While proper development is not rocket science, it does take somebody who knows what they are doing, and not guessing, to get it right.