Checking over the logs it became obvious that queries using SQL_CALC_FOUND_ROWS are a problem. Digging through code it appears that where a limit on results is used it uses this query modifier to allow simpler coding of paging. The problem is that since the main query to grab blog posts on the blog home page or any archive page uses this, on a large blog this becomes quite inefficient.

For now I have tweaked MySQL so that the MySQL Query Cache is now larger. These articles were useful:
MySQL Query Cache (at MySQL Performance Blog)
Turn on MySQL query cache to speed up query performance?

Looking at the log after this change query times have dropped dramatically (from over 0.5 seconds and up to milliseconds). This has immediately improved performance of the blog and overall load on the server.

I am currently looking at some WP Caching API and plugins to further improve performance.

If you are looking for some more ways to speed up WordPress check out this article as a starting point:
4+1 Ways To Speed Up WordPress With Caching

And … for myself – just so I remember:
Install Memcache onto cPanel running CentOS

Real World Web: Performance & Scalability:

Here is a good rundown of setting up a VAMP (Vista, Apache, MySQL, PHP) server from Devshed:
Creating a VAMP

The other thing to know, which this article skipped, is that you should run your text editor as administrator when editing the config files. If you don’t do this you won’t be editing the correct file (only your Users version of it which isn’t actually used by the system).

I have just been reading this …
ORDER BY … LIMIT Performance Optimization. Pagination is used on just about every site I work on – that article is great on speeding up queries using ORDER BY and LIMIT.

The only thing to configure is the database host, user, and password. Simply upload the dump file to the same directory as the bigdump file (use FTP or the built-in uploader) and then process that file.