Us geeks usually implement clusters because the business wants higher availability. Higher availability means faster troubleshooting when the system is down. We need to be able to get the system back up and running as quickly as possible. Getting there usually means reducing the amount of complexity; complex systems take longer to troubleshoot.
Like you here all the time, Keep it simple stupid, or the nicer way, keep it stupid simple.