The simple fact is that in the case of search engines as opposed to directories (e.g. DMOZ and Yahoo), the site is not read by a human, necessarily, before being listed: it is read by a robot or spider.
This means that:
Spiders are also a little lazy. They don't like searching through a lot of content, so I'm lead to believe, and it's certainly worth ensuring a decent summary of your page's content comes first in the document, and isn't held up by uneccessary content.
By uneccessary content, I mean:
before the first thing in the body of your page, which should be the page's title, wrapped, semantically, in <h1> tags.
In terms of good website design, i.e. keeping the styling and functionality seperate from the mark up, styling and scripts should be kept in seperate files anyway. It also means your lazy spider only has the one line to read through before all the javaScript or CSS for the page is done (read: ignored).
Comments are obviously helpful and are read by spiders, but if the spider you reading your site as a limit on the number of characters it can read, it's not really worth filling them with commented out code high in the file.
Meta tags used to be thought of as the silver bullet for search engines, before they were spammed full of 'SEX SEX SEX', thus leaving them with less weight. It's still worth futureproofing your page, and allowing for those spiders that do read them with a few space-seperated keywords and a pithy description, which, of course, should actually describe the page. This is just good practice.