We use clustering in production and log ship to a secondary site to maintain a standby server, this gives us the best of both worlds in my opinion.

To be honest I don't like the idea for transactional replication for standby servers, not when there are other options available, but that's only my opinion and i'm sure that others will disagree.