this being my first read of the article, the following caught my attention:

"...generally I recommend log-shipping for fail-over, followed by transactional replication with queued updating subscribers. The order is reversed if there is a possibility of needing to go back to the production server ..."

The question came to mind whether or not Paul was indicating that the two different methods could be used in combination somehow, or was he simply stating that his first preference is log shipping and his second preference is transactional replication.

Can anyone clarify this for me?

If he was indicating his preferences in order, I wonder, subsequently, if it is possible to combine the methods for a sort of poor-man's version (e.g. no 3rd party utilities) of the best of both worlds? Or would this be the SQL equivalent to smoking crack?