For example, if you search Google for my blog, with a keyword of “hawking”, you get the following results.
Most of them point to the right place at some point in time, but they point to many different places over time. Because blogs evolve. Content moves from the front page to the second page, from the main blog to the archives.
Google Search could easily determine the pattern of page migration of my blog, and 90% of the other blogs on the planet – and always use the most relevant link. In this instance the only really relevant links are to the archive of the Hawking post, and to comments on the post. Not to where it used to be on the front page. Not to where it used to be on the second page, etc. You won’t find the post on those pages anymore. But Google doesn’t forget – at least not quickly enough.
Properly indexing blogs is something I think Google should have already tackled by now. It’s not rocket science, after all.