When building a Drupal 6 site (mysql+php5), are there any reasons not to use Pressflow on the production server?
There is nothing that should explicitly stop you from doing so. However there are some things that may be worth considering.
There are less developers working on pressflow and there may be a loger time to get core bugs fixed. Also there are less people using it so more chance of finding an undiscovered bug.
While pressflow is API compatible, most modules are not tested against pressflow so a particular module may not work with pressflow.
Security fixes may take longer to get into pressflow than Drupal core.
It is potentially more difficult to configure as there are settings aimed at high end performance.
Finally I suppose I would turn round the question, unless you have a website with a lot of load, why would you use a distribution aimed at high traffic sites. A lot of the scalability gains in pressflow have been back ported from D7, so you may consider that as well.
Note, I don’t have anything against pressflow, I have used it and it has worked well and the people who work on it are very clever. I am just pointing out general risks with using a less popular branch.